Language：English   Publishing type：Research paper (international conference proceedings)  

DOI： 10.1117/12.2521497

Scopus

researchmap

Authorship：Lead author   Language：English   Publishing type：Research paper (scientific journal)  

DOI： 10.1109/TBC.2018.2829118

Web of Science

Scopus

researchmap

Publisher：IEEE  

DOI： 10.1109/BMSB.2018.8436914

researchmap

Language：English   Publishing type：Research paper (international conference proceedings)  

DOI： 10.1109/GCCE.2017.8229189

Web of Science

Scopus

researchmap

Language：English   Publishing type：Research paper (international conference proceedings)  

Web of Science

researchmap

Language：English   Publishing type：Research paper (international conference proceedings)  

DOI： 10.1109/BMSB.2016.7521902

Web of Science

researchmap

Publishing type：Research paper (scientific journal)  

DOI： 10.3169/itej.69.J23

Scopus

researchmap

Language：Japanese   Publishing type：Research paper (scientific journal)  

researchmap

Language：English   Publishing type：Research paper (international conference proceedings)   Publisher：Institute of Electrical and Electronics Engineers Inc.  

DOI： 10.1109/ICCE.2014.6775906

Scopus

researchmap

Language：English   Publishing type：Research paper (international conference proceedings)  

Web of Science

researchmap

Authorship：Lead author,　Corresponding author   Language：English   Publishing type：Research paper (scientific journal)  

DOI： 10.1109/JPROC.2012.2205656

Web of Science

researchmap

Publishing type：Research paper (scientific journal)   Publisher：Institute of Image Information and Television Engineers  

DOI： 10.3169/mta.1.27

researchmap

Publishing type：Research paper (scientific journal)  

Scopus

researchmap

Language：English   Publishing type：Research paper (international conference proceedings)   Publisher：Society of Motion Picture and Television Engineers  

Scopus

researchmap

Language：English   Publishing type：Research paper (international conference proceedings)   Publisher：Blackwell Publishing Ltd  

DOI： 10.1002/sdtp.51

Scopus

researchmap

Language：English  

DOI： 10.1109/JPROC.2011.2182075

Web of Science

researchmap

Language：English   Publishing type：Research paper (international conference proceedings)  

Web of Science

researchmap

Language：English   Publishing type：Research paper (international conference proceedings)  

Scopus

researchmap

Language：English   Publishing type：Research paper (international conference proceedings)  

Scopus

researchmap

Language：English   Publishing type：Research paper (international conference proceedings)  

Web of Science

researchmap

Language：English   Publishing type：Research paper (international conference proceedings)  

DOI： 10.1109/PCS.2012.6213357

Web of Science

researchmap

Language：English   Publishing type：Research paper (international conference proceedings)  

DOI： 10.1109/GCCE.2012.6379908

Scopus

researchmap

Language：Japanese   Publisher：一般社団法人電子情報通信学会  

映像符号化におけるイントラ予測の新しい方式を提案する.この方式は,反復処理を用いることで隣接領域との連続性を維持しながら符号化対象ブロックのテクスチャを予測する.実験により提案方式がAVC/H.264方式で使用されている方式より高い予測性能をもつことを示す.

CiNii Research

researchmap

Language：English   Publishing type：Research paper (international conference proceedings)  

DOI： 10.1109/ICIP.2011.6115634

Web of Science

researchmap

Language：English   Publishing type：Research paper (international conference proceedings)  

Web of Science

researchmap

Language：Japanese   Publisher：一般社団法人電子情報通信学会  

超高精細映像は,時間標本化周波数に比べて高い空間標本化周波数をもつ.ゆえに超高精細映像は,従来の映像に比べて大きな動き量と動領域面積を示す可能性が高い.このとき動領域ぼけは大きくなるため,大きなブロックや探索範囲を用いる高確度レジストレーション手法が有効である.ただし超高精細映像においては計算量の増大を招くため,本論文では画像ごとに時空間高周波パワーを解析して画像がもつ動きぼけや動領域面積を推定し,結果に応じて階層的にレジストレーションを行う手法を検討する. 2.では,空間,時空間高周波帯域がもつパワーと動き量を解析する.解析及び考察の結果,超高精細映像の空間,時空間高周波帯域がもつパワーの割合と,動き量を関連づける有効なデータが得られた.結果より,動き量と動領域面積の大きな画像では,時間周波数帯域全体に対する高周波帯域がもつパワーの割合は高く,空間周波数帯域全体に対する高周波帯域がもつパワーの割合は低い. 4.では,上記解析結果を踏まえて,時空間ウェーブレット解析を用いた階層型レジストレーション手法の一検討を行う.提案法は,時間高周波帯域がもつパワーの割合を解析し,高い場合は空間分解階数を大きな値に設定して階層型レジストレーションを行うことを特徴とする.各階層では同じブロックサイズ,動き探索範囲を用いるため,空間分解階数が大きい階層では,大きなブロックサイズと動き探索範囲を用いることに相当する.

CiNii Research

researchmap

Language：English   Publishing type：Research paper (scientific journal)  

DOI： 10.5594/J12203

Web of Science

researchmap

Language：Japanese   Publisher：The Institute of Image Information and Television Engineers  

We propose a novel intra prediction method. It applies a discrete cosine transform and an inversely transform to an image region repeatedly to obtains better intra prediction pixel values.

DOI： 10.11485/itewac.2010.0_3_1

CiNii Research

researchmap

Language：English   Publishing type：Research paper (international conference proceedings)   Publisher：IEEE  

DOI： 10.1109/PCS.2010.5702537

Scopus

researchmap

Language：English   Publishing type：Research paper (scientific journal)   Publisher：Society of Motion Picture and Television Engineers  

DOI： 10.5594/J12203

Scopus

researchmap

Language：English   Publishing type：Research paper (international conference proceedings)   Publisher：ACM  

DOI： 10.1145/1873951.1874074

Scopus

researchmap

Other Link： http://dblp.uni-trier.de/db/conf/mm/mm2010.html#conf/mm/MisuMSS10

Language：English   Publishing type：Research paper (international conference proceedings)   Publisher：IEEE  

DOI： 10.1109/PCS.2010.5702498

Scopus

researchmap

Language：Japanese   Publisher：The Institute of Image Information and Television Engineers  

This paper proposes a novel iterative super-resolution method that generates high-frequency information for an enlarged image. The block-wise high-frequency information is obtained through a simulated process of down-scaling and up-scaling of a corresponding larger homothetic block. The experimental results show the effectiveness of the proposed

DOI： 10.11485/itewac.2010.0_4_7

CiNii Research

researchmap

Language：English   Publishing type：Research paper (international conference proceedings)  

DOI： 10.1109/ICCE.2010.5418823

Web of Science

researchmap

Language：Japanese   Publisher：情報処理学会  

近年，放送のデジタル化と多チャネル化により日々大量のニュースが発信されるようになった．また，視聴者の嗜好や視聴形態は多様化しているため，関心のあるニュースや必要な情報だけを短時間で視聴するといった個人に適合した視聴方法に対する要望が高まってきた．このような要望に応えるため，筆者らは受信機側で複数のニュース番組から必要なストーリのみを抽出して適切な順序に並べるダイジェスト作成システムを構築している．従来のダイジェスト作成システムでは，ユーザの嗜好をより正確に学習することや，よりユーザの嗜好に合ったストーリを抽出することなどが主なテーマであった．しかし，ユーザの嗜好に合ったニュースを提示するだけでは，ユーザは現在話題となっている情報や緊急情報などの重要なニュースを見逃してしまう恐れがある．そこで本研究では，ニュースダイジェストを作成するにあたりニュースの重要性を考慮できるよう，ニュースの送出状況により受信機側でニュースの重要度を算出する手法を提案する．また，実際に放送局のニュース編集責任者としての経験を有する者が判定する重要度と比較することで，本手法の妥当性を評価する．In recent years, due to the digitalisation of broadcast and the increase of broadcast channels, the number of news delivered to users has multiplied more and more. Additionally, the user's preference and their demand for specific ways to view news programs, such as short-viewing of only necessary news, have been diversifying. Therefore, authors are constructing a news digest system extracting only necessary news stories from some news programs and laying out them in the order of their necessity. The previous works of creating digests have mainly dealt with estimating the user's preference accurately, and extracting the stories adapting to the user's preference correctly. However, if the users are provided only the news stories adapting to their preferences, they might miss the important information such as hot topics and emergency news. Consequently, we propose a method of calculating news value using broadcasting states of news stories for creating news digests. This paper also evaluates the method by comparing the values which the system calculates to those which the executive editors of news programs judge.

CiNii Research

researchmap

Language：English   Publishing type：Research paper (international conference proceedings)  

Web of Science

researchmap

Language：Japanese   Publisher：The Institute of Image Information and Television Engineers  

A new graphical user interface is proposed. With proposal interface, user can select information from a number of options using a remote controller with a trackball viewing on TV display. Information options are allayed in lines in 3-dimensional space and displaying size and position of each option are changed according to user's control.

DOI： 10.11485/iteac.2009.0__5-8-1_

CiNii Research

researchmap

Language：English   Publishing type：Research paper (international conference proceedings)  

DOI： 10.1109/GLOCOM.2008.ECP.333

Web of Science

researchmap

Language：Japanese   Publisher：一般社団法人電子情報通信学会  

本論文では,ネットワークを利用した分散処理環境下において,任意の機器間でコンテンツの相互運用性を実現するためのオブジェクトモデルを提案する.提案するオブジェクトモデルは,システムをメディアオブジェクトと機能モジュールと呼ぶ2種類のオブジェクトで構成する.ユーザからの要求に対して仲介役として機能するブローカモジュールが,必要な機能モジュールの探索,経路決定,接続要求を行う.これにより,コンテンツの信号形式にかかわらず,任意の機器でコンテンツを扱うことが可能となる.提案するオブジェクトモデルの有効性を確認するための検証用システムとして,映像再生システムを試作した.このシステムでは,コンテンツの信号形式にかかわらず,コンテンツの再生に必要な機器がネットワーク上で自動的に接続される.この試作システムにより,ユーザがコンテンツの信号形式を意識することなく再生を行うことができるなど,提案するオブジェクトモデルの有効性と,実現の可能性が確認できた.

CiNii Research

researchmap

Language：Japanese   Publisher：一般社団法人電子情報通信学会  

本論文では,高品質なストリーミングサービスの提供を目指して,伝送路におけるパケットロス要因となるバーストトラヒックをストリーミングサーバ自体で抑制する配信制御方式を提案し,ソフトウェアによる実装方法を示す,本方式は,同時送出するすべてのストリームの符号化レートやパケット長を考慮して,個々のパケットの送出タイミングと出力順序を一意に決定することにより,個別ストリームの突発的な連続出力,及び出力ストリーム全体としての瞬間的なビットレート上昇を同時に抑制することを特徴とする.また,送出スレッドで一括してタイミング制御を行うこと,及び処理遅延を考慮した制御マージンを設定することにより精度の高いタイミング制御をソフトウェアにより実現する.提案方式をソフトウェアにより実装しバースト抑制性能の評価を行った結果,汎用OS上の実装であっても十分な性能を示すことが確認できた.更に実験結果をもとに,ネットワークノードにおけるバッファ滞留量の低減効果を確認し,提案方式の有効性を示した.

CiNii Research

researchmap

Language：Japanese   Publisher：The Institute of Image Information and Television Engineers  

The spread of broadband IP networks is making high-quality video distribution service a reality. However, some technical problems should be solved to deliver programs nationwide with quality and reliability equivalent to that of current broadcast services using radio wave. This study illustrates high level requirements of realizing technology of high quality and dependable IP broadcasting service.

DOI： 10.11485/iteac.2005.0__17-5-1_

CiNii Research

researchmap

Language：Japanese   Publisher：The Institute of Image Information and Television Engineers  

Streaming contents of TV programs over IP network may not be received just as broadcasting corporation intended. It is a result of network trouble like packet losses. This is one of the problems to be solved to realize TV broadcasting over IP network. This paper describes an experimental system which keeps broadcasting the information of programs even if network trouble occurs.

DOI： 10.11485/iteac.2005.0__17-6-1_

CiNii Research

researchmap

Language：Japanese   Publisher：The Institute of Image Information and Television Engineers  

Broadband IP network is gradually making feasible video distribution services. Studying how to transmit TV programs if 1P network trouble like packet losses occurs is important. We propose a system which controls an amount of adding redundancy depending on the extent of network trouble to transmit at least what TV programs say. In this paper we make an estimate of an amount of adding redundancy to ensure target quality at a receiver.

DOI： 10.11485/iteac.2004.0__4-6-1_

CiNii Research

researchmap

Language：English   Publishing type：Research paper (scientific journal)  

DOI： 10.1002/scj.10292

Scopus

researchmap

Language：Japanese   Publisher：The Institute of Image Information and Television Engineers  

A new TV navigation system to support viewing TV programs was developed and subjective evaluation tests were carried out. It recommends optimum programs depending on viewing habit and viewers profile besides viewers' easy program selection.

DOI： 10.11485/itetr.27.14.0_9

CiNii Research

researchmap

Language：English   Publishing type：Research paper (scientific journal)  

DOI： 10.1109/TCSVT.2002.1003471

Web of Science

researchmap

Language：English   Publishing type：Research paper (scientific journal)  

DOI： 10.1109/TIP.2002.999672

Web of Science

researchmap

Language：Japanese   Publisher：一般社団法人電子情報通信学会  

本論文は,画像の局所的な特徴を考慮した固有値こう配法による動き推定手法を提案する.提案する手法は,はじめに固有値こう配法による動き推定を行い,信頼度の低い推定箇所を推定に用いたブロック内の画像の特徴から分類する.信頼度の低い推定箇所の推定精度を改善するために,本手法では二つの固有値こう配法に基づく手法を提案する.複数の動きが含まれると分類されたブロックに対しては,ロバスト推定を適用した固有値こう配法による動き推定を行う.一方,ブロック内に1種類のこう配のみ存在すると分類されたブロックに対しては,エッジを共有する画像領域の他の位置のブロックの固有ベクトルを,推定対象の固有ベクトルとして補完することで動き推定を行う.実験の結果,提案手法は少ない計算コストの増加で,動き推定の精度を改善できることを確認した.

CiNii Research

researchmap

Language：Japanese   Publisher：一般社団法人電子情報通信学会  

動画像符号化において, ブロック分割に基づく動き補償は必須の技術として用いられている.動き補償フレーム間差分信号について, ブロック境界部の補償差分の分散が中心部より大きいという空間的非均一性が観測されているが, この特性に関する理論的解析はこれまでに行われていない.本論文では, まず統計的動き分布モデルを導入することによりこの特性を理論的に解明する.そして, このモデルをオーバラップブロック動き補償の性能解析に応用し, オーバラップブロック動き補償がブロックひずみの軽減のみならず, 補償差分信号の低減にも有効であることを理論的に明らかにする.更に, 理論解析の結果を実験より確認する.

CiNii Research

researchmap

Language：Japanese   Publisher：一般社団法人電子情報通信学会  

本論文は, こう配法に基づく動き推定において, ブロック内の画像特徴に応じて適応的にブロックサイズを設定する手法について述べる. 提案手法は, ブロック内の輝度こう配ベクトルの分布状態から誤推定を引き起こす要因を判別し, ブロックサイズの修正処理を行う.

CiNii Research

researchmap

Language：English   Publishing type：Research paper (scientific journal)  

DOI： 10.1109/11.920780

Web of Science

researchmap

Language：English   Publishing type：Research paper (international conference proceedings)  

DOI： 10.1109/ICIP.2001.958166

Web of Science

researchmap

Language：Japanese   Publisher：The Institute of Image Information and Television Engineers  

This paper describes subjective evaluation results of low bit-rate video coding. The evaluation tests are based on the terms which are derived from the SD (Semantic Difference) and the factor analysis. The obtained data will be helpful to the design of MPEG-4 video encoding system in the digital terrestrial television broadcasting.

DOI： 10.11485/iteac.2001.0_47

CiNii Research

researchmap

Language：English   Publishing type：Research paper (scientific journal)   Publisher：Scripta Technica Inc  

DOI： 10.1002/(SICI)1520-684X(200007)31:7<10::AID-SCJ2>3.0.CO;2-U

Scopus

researchmap

Language：English   Publisher：The Institute of Image Information and Television Engineers  

Information about camera operation such as zooming, focusing, panning, tilting and dollying is useful not only for efficient video coding, but also for detection of moving objects in the scene. We have developed a camera operation parameter measurement system and applied it to inferring the camera-induced image motion directly from pan/tilt/zoom parameters. However, when the camera is undergoing translational movement such as dollying, the camera-induced 2D motion depends both on the camera operation parameters and the depth of each point in the scene. In this paper, we propose a multiframe image matching algorithm for robust depth-map estimation. The algorithm exploits camera operation parameters to establish point correspondence among multiple image frames. The experimental results show significant improvement in the accuracy of depth-map compared with the conventional two-frame matching method.

DOI： 10.11485/itetr.24.2.0_85

CiNii Research

researchmap

Language：Japanese   Publisher：The Institute of Image Information and Television Engineers  

This paper proposes an inter frame lossless shape image coding method based on the contour direction vector expression. Motion vectors are detected by the matching method for the contour direction vectors. After wavelet transform, the contour direction vectors are encoded by the arithmetic coding. The simulation result shows that the proposed method is more efficient than that of MPEG-4.

DOI： 10.11485/iteac.2000.0_66

CiNii Research

researchmap

Language：English   Publishing type：Research paper (international conference proceedings)  

DOI： 10.1109/ICIP.2000.899821

Web of Science

researchmap

Language：English   Publishing type：Research paper (scientific journal)  

DOI： 10.1109/TIP.2002.999672

Scopus

researchmap

Language：Japanese   Publisher：The Institute of Image Information and Television Engineers  

This paper describes a motion estimation method with eigenvalue algorithm matching local image feature. The proposed method first formulates the optical flow using the block gradient method with eigenvalue algorithm, and then low reliability optical flow is classified according to the characteristics inside the block. We introduce two estimation methods based on eigenvalue algorithm for improvement in accuracy. The first method is based on robust estimation. We apply this method to the blocks which include multiple movements. The second method is applied to the blocks in which one typical gradient alone exists. In the method, we complement the eigenvector from the object with the same edge in the subjective block, and then motion estimation is carried out with the complementary eigenvector. Experimental results show that the proposed method improves the estimation accuracy.

DOI： 10.11485/itetr.23.80.0_37

CiNii Research

researchmap

Language：Japanese   Publisher：一般社団法人電子情報通信学会  

MPEG-2映像符号化規格の範囲内での符号化画質の改善研究が行われている. 本論文では, MPEG-2符号化においてI, P, Bのピクチャタイプへ符号化に要する符号量を適切に配分する符号量配分手法を提案する. これまで, GOP当りのひずみ量を推定し, ビットレートの拘束条件の下でラグランジュ未定係数法を適用することで配分符号量を求める手法が提案されている. しかし, その手法で推定したひずみ量は, 三つのピクチャタイプの信号特性が十分に組み込まれているとはいえない. そこで, ピクチャタイプの信号特性を考慮したGOP当りのひずみ量を推定し, 最適な符号量配分を行う手法を提案する. 本手法で推定するGOP当りのひずみ量の特徴は, ビクチャタイプ別のレートひずみ特性, 及び予測参照画像のひずみ成分が当該符号化画像の符号化効率に及ぼす影響を考慮する点である. MP◎HL 18Mbit/sの符号化シミュレーションによりMPEG-2のTM5との比較を行った結果, 3種類の画像に対して符号化SNRで0.4dB程度の改善が得られた. また, 本手法がほぼ最適な符号量配分を行っていることを示す.

CiNii Research

researchmap

Language：English   Publisher：The Institute of Image Information and Television Engineers  

Information about camera operating such as zoom, focus, pan, tilt and tracking is useful not only for efficient video coding, but also for content-based video representation. A camera operating parameter measurement system designed specifically for these applications is therefore developed. This system, implemented in real time and synchronized with the video signal, measures the precise camera operating parameters. We calibrated the camera lens using a camera model that accounts for radial lens distortion. The system is then applied to infer image motion from pan and tilt operating parameters. The experimental results show that the inferred motion coincides with the actual motion very well, with an error of less than 0.5 pixel even for large motion up to 80 pixels.

DOI： 10.11485/itetr.23.3.0_47

CiNii Research

researchmap

Publisher：IEEE  

DOI： 10.1109/ICIP.1999.817205

researchmap

Language：English   Publishing type：Research paper (international conference proceedings)  

Web of Science

J-GLOBAL

researchmap

Language：English   Publishing type：Research paper (international conference proceedings)  

Scopus

researchmap

Language：Japanese   Publisher：The Institute of Image Information and Television Engineers  

This paper describes a method of computational complexity reduction for the gradient motion estimation with eigenvalue algorithm. The proposed method adopts the algorithm based on perturbation theorem. The experimental results have proved that the proposed method can reduce the computational complexity and estimate movement with the same accuracy as the conventional algorithm.

DOI： 10.11485/iteac.1999.0_308

CiNii Research

researchmap

Language：Japanese   Publisher：一般社団法人電子情報通信学会  

動画像領域分割符号化においては, 画像内の各画素について正確な動き推定を行うことが重要である.従来のブロックマッチング手法では, 推定した動き量が被写体の動きを必ずしも正確に反映しないという推定誤りが生じることがある.空間解像度の高い動きベクトル場を求めるためには, ブロックサイズを小さくする必要があるが, その場合ランダムな推定誤りの発生が多くなる.本論文では, 推定誤りの低減, 動きベクトル場の空間解像度を上げることを同時に可能にするために, 動き量の時間方向の連続性に着目し, 複数フレームを用いた階層化時空間3次元ブロックマッチング手法を提案する.提案手法は, 従来の1フレーム手法に比べ, ランダムな推定誤りを増やすことなく空間ブロックサイズを小さくすることができ, 空間解像度の高い動きベクトル場の推定を行うことが可能となる.

CiNii Research

researchmap

Language：Japanese   Publisher：一般社団法人電子情報通信学会  

画像のオブジェクトベース符号化を行うためには, 前処理として領域分割が必要である.本論文では領域の統合処理を伴う実用的な手法を提案する.画像の領域分割法として広く利用されているK平均アルゴリズムによるクラスタリングは, 画像を小さな領域に過分括する傾向がある.また, 最初に与えるクラスタの形状に影響され本来の領域の境界以外の部分で分割される, あるいは初期クラスタ数の増加に応じて得られる領域の数が増大するという初期値依存性を有する.これらの問題点に対処するために, 本論文ではK平均アルゴリズムをベースとして, 多数決フィルタリング, ラベリングによる統合, 複数の分割結果の統合の3段階の処理手法を導入する.この手法を数種類の自然画像に適用し, 同一の領域分割処理のパラメータを用いても, 異なる画像中の物体の輪郭を抽出できることを確認した.また, 初期に与えるクラスタの数を変化させても領域を同様に抽出できることも確認した.更に複数の画像を用いたシミュレーション実験によりその有効性を検証した.本手法は少ない演算量で良好な領域分割結果が得られるだけでなく, 初期クラスタや画像に影響されないロバストな処理であるため, オブジェクトベース符号化の要素技術として有効である.

CiNii Research

researchmap

Language：English   Publisher：The Institute of Image Information and Television Engineers  

This paper describes the concept of object-based coding for future broadcasting environments. Digital broadcasting uses the MPEG2 coding scheme which is regarded as a picture-based coding. An object-based coding scheme is a potential candidate for future broadcasting both for studio and distribution uses, and it offers a higher compression and more flexible content handling. This paper also describes key technologies that we have been developing for the object-based coding, e.g., image analysis, object extraction and coding of objects.

DOI： 10.11485/itetr.22.2.0_25

CiNii Research

researchmap

Publishing type：Research paper (scientific journal)   Publisher：NHK Laboratories Note  

Scopus

researchmap

Language：English   Publishing type：Research paper (scientific journal)   Publisher：John Wiley and Sons Inc.  

DOI： 10.1002/(SICI)1520-684X(199812)29:14<68::AID-SCJ8>3.0.CO;2-P

Scopus

researchmap

Publishing type：Research paper (scientific journal)  

Scopus

researchmap

Language：Japanese   Publisher：一般社団法人電子情報通信学会  

現在, オブジェクト単位あるいは領域単位の符号化が活発に検討されている. このような方式において方形でない形状をもつ画素の集まりをどのように符号化するかが重要な要素となる. 今回, この符号化法として領域サポートDCTを提案する. 提案手法は従来の2次元DCTを基本としており, 復号処理は従来のものをそのまま使えるという特長を有する. また, RD特性もSA-DCT等の既に提案されている任意形状符号化手法と同等以上の性能を示している.

CiNii Research

researchmap

Language：Japanese   Publisher：一般社団法人電子情報通信学会  

本論文では, 立体画像符号化への応用を前提として, 階層的ブロック相互相関視差推定手法を提案し, その性能を検証する. 性能検証に際し, 従来の補償誤差電力による評価では不十分なため, まず符号化誤差による新しい評価法を提案する. 次に, この新評価法を用いて, SADによる全探索ブロックマッチング, SAD による階層的ブロックマッチング, 階層的ブロック相互相関法の三つの視差推定手法に対し性能比較を行う. 実験結果より, 階層的ブロック相互相関法は他の手法に比べ, 視差ベクトルの差分エントロピーが10〜30%の減少となること, 補償誤差の符号化誤差については同等以上の性能をもつことがわかった. また, 推定された視差の信頼性がより高いので, 低ビットレート伝送, 画像内挿などへ応用する場合は, 破たんの少ない, より良い画質の画像が得られると考えられる.

CiNii Research

researchmap

Publishing type：Research paper (scientific journal)  

DOI： 10.1002/(sici)1520-684x(19970615)28:6<1::aid-scj1>3.0.co;2-f

Scopus

researchmap

Language：English   Publishing type：Research paper (international conference proceedings)  

Scopus

researchmap

Language：English   Publishing type：Research paper (scientific journal)  

Web of Science

researchmap

Language：English   Publishing type：Research paper (scientific journal)  

Web of Science

researchmap

Language：Japanese   Publisher：一般社団法人映像情報メディア学会  

In this paper, we will examine, through subjective assessment tests, the effectiveness of the bi-directional prediction scheme, which was adopted by MPEG as a standardized coding technique and which is generally assumed to be effective with respect to the coding efficiency. We will also clarify the distinctive picture quality degradation by way of the bi-directional prediction coding scheme, analyze causes, and evaluate the de7gradation quantitatively. Specifically, the assessment results show that the bi-directional prediction scheme works well in most of the sequences with respect to the picture quality improvement, but causes flickers to some sequences. The analyses show that for the picture with high contrast texture the resolution of the decoded image differs between the B-pictures and the P-pictures; this leads to the flicker through the gamma characteristics of a monitor. We will demonstrate that we can evaluate the flicker by simulating the electro-optic conversion process.

DOI： 10.3169/itej1978.50.391

CiNii Research

researchmap

Language：Japanese   Publisher：The Institute of Image Information and Television Engineers  

We previously proposed RS(Region Support)-DCT, a novel coding tool for texture within an arbitrarily shaped region. In this paper, we study a derivation process of a RS-DCT coefficient set. The result shows that if a conventional type VLC method is assumed, an introduction of a frequency weighting into the coefficient derivation works well.

DOI： 10.11485/tvac.32.0_327

CiNii Research

researchmap

Language：Japanese   Publisher：一般社団法人電子情報通信学会  

インタレース走査されたテレビ信号の高能率符号化において,近年,二つのフィールドを組み合わせてフレーム信号(フィールドマージ信号)とし,これに対し空間的な符号化を行う手法も用いられている.符号化にとって信号源の電力スペクトル構造は重要な基礎であるが,フィールドマージ信号のスペクトル構造はあまり明らかにされていない.また,現在用いられている符号化手法の最適性も不明である.そこで,フィールドマージ信号の電力スペクトルモデルを理論的に導出し,その性質を動きの大きさとの関連において明らかにした.また,これを用い,フィールド単位処理との符号化効率の比較を行い,フレーム単位処理の導入の効果を試算した.

CiNii Research

researchmap

Language：English   Publishing type：Research paper (international conference proceedings)  

Web of Science

researchmap

Language：Japanese   Publisher：一般社団法人電子情報通信学会  

ハイビジョン信号はその広い帯域のため信号に含まれる雑音が多く,符号化効率の低減の原因となる.しかし,素材伝送のような高忠実度伝送においては前処理により雑音除去を行うことは望ましくない.そこで,予測ループ内の信号の雑音成分を除去し,予測効率を上げるためループフィルタの導入を検討した.この中で,その望ましい特性の導出も行った.また,動き補償精度,量子化特性,ハイビジョンと標準テレビの違いに着目し,種々の条件下でシミュレーションを行い,ループフィルタの有効性・有効範囲の検証を行った.その結果,提案のループフィルタは素材伝送用途の符号化には非常に有効であること,特に1画素精度の動き補償を用いるときは必須のものであることがわかった.

CiNii Research

researchmap

Language：Japanese   Publisher：一般社団法人電子情報通信学会  

アナログ帯域圧縮法としてテレビジョンシステムで用いられているインタレースの得失をディジタル伝送の観点から検討するため,電力スペクトルモデルを用いて順次走査信号とインタレース信号のレートひずみ特性の理論的な比較を行った.

CiNii Research

researchmap

Publishing type：Research paper (scientific journal)  

DOI： 10.1002/ecja.4410760806

Scopus

researchmap

Language：Japanese   Publisher：電子情報通信学会通信ソサイエティ  

CiNii Research

researchmap

Language：English   Publishing type：Research paper (scientific journal)  

Web of Science

J-GLOBAL

researchmap

Language：English   Publishing type：Research paper (scientific journal)  

DOI： 10.1016/0923-5965(92)90010-D

Scopus

researchmap

Language：Japanese   Publisher：The Institute of Image Information and Television Engineers  

Studies have been conducted for the development of an HDTV coding and transmission system for contribution purposes. This paper describes the results of investigations into loop filters used in the coding systems. HDTV signals contain a large amount of noise due to their wide bandwidth. This noise leads to poor coding efficiency. A pre-processing of coding input signals, however, is not preferable for high-fidelity transmission. Hence, the introduction of a loop-filter is examined, which reduces the noise in the prediction loop and increases the prediction accuracy. An investigation into a suitable filter characteristics is also conducted. The results show that the loop-filter is very effective for contribution-purpose coding systems, especially for ones using one pel accuracy motion compensation.

DOI： 10.11485/tvtr.16.77_1

CiNii Research

researchmap

Language：Japanese   Publisher：一般社団法人映像情報メディア学会  

Motion compensated prediction is one of the most important techniques in image coding. A comparison between motion compensated interfield and interframe predictions is conducted through theoretical investigation and computer simulation. The results show that interfield prediction has advantages of better or equal performance to interframe prediction and narrower vector search range, although it is disadvantageous for certain kinds of movement, such as horizontal panning. The results also indicate that hybrid motion compensation, where interfield motion compensation is combined with horizontally one-directional interframe motion compensation, is an effective method.

DOI： 10.3169/itej1978.46.84

CiNii Research

researchmap

Language：Japanese   Publisher：The Institute of Image Information and Television Engineers  

A study on a coding scheme which is appropriate for motion compensation prediction (MC) error signals is presented. MC error signals have very small correlation between adjacent pixels, especially in the vertical direction. Moreover, the correlation varies to a large extent depending on scene, local feature and type of movement. Therefore, with respect to coding gain and decoded picture quality, an adaptive coding scheme where orthogonal transform or subband partition is applied only for cases having high correlation is likely efficient for MC error signals.

DOI： 10.11485/tvtr.16.11_19

CiNii Research

researchmap

Language：Japanese   Publisher：The Institute of Image Information and Television Engineers  

An HDTV codec based on DCT plus motion-compensated prediction at a bitrate of 135 Mb/s is now being developed. The hardware prototype will be completed soon. This paper describes the specification and the configuration of the codec.

DOI： 10.11485/tvtr.15.32_1

CiNii Research

researchmap

Language：English  

J-GLOBAL

researchmap

Language：Japanese   Publisher：The Institute of Image Information and Television Engineers  

DOI： 10.11485/tvtr.14.28_25

CiNii Research

researchmap

Language：English   Publishing type：Research paper (international conference proceedings)  

Web of Science

researchmap

Language：Japanese   Publisher：The Institute of Image Information and Television Engineers  

DOI： 10.11485/ncrtv.24.0_191

CiNii Research

researchmap

Language：English   Publishing type：Research paper (scientific journal)  

Web of Science

researchmap

Language：English   Publishing type：Research paper (scientific journal)  

Web of Science

J-GLOBAL

researchmap

Language：English  

J-GLOBAL

researchmap

Language：Japanese  

CiNii Research

researchmap

Language：English   Publishing type：Research paper (scientific journal)  

DOI： 10.1002/ecja.4410650403

Scopus

researchmap

Language：Japanese   Publisher：The Institute of Image Information and Television Engineers  

DOI： 10.11485/ncrtv.17.0_177

CiNii Research

researchmap

Language：Japanese   Publisher：The Institute of Electronics, Information and Communication Engineers  

CiNii Research

researchmap

Language：Japanese   Publisher：日本放送協会放送技術研究所  

CiNii Research

researchmap

Language：Japanese   Publisher：The Institute of Image Information and Television Engineers  

NHKは次世代の放送サービスとしてスーパーハイビジョンの研究開発を進めている.ロンドンオリンピックでスーパーハイビジョンのパブリックビューイングを日本,イギリス,アメリカの3ヵ国9会場において実施し,多くの方から大変高い評価をいただいた.ここでは,このパブリックビューイングの概要について紹介する.

DOI： 10.3169/itej.66.1023

CiNii Research

researchmap

Language：Japanese   Publisher：日本放送協会放送技術研究所  

CiNii Research

researchmap

Language：Japanese   Publisher：Forum on Information Technology  

CiNii Research

researchmap

Language：Japanese   Publisher：The Institute of Electronics, Information and Communication Engineers  

The Super Hi-Vision (SHV) is an ultra high-definition video system with 4,000 scanning lines. Its video format is 7,680 x 4,320 pixels and the frame rate is 60 Hz with progressive scanning. It has been designed to give viewers a strong sense of reality. To make the system suitable for practical use, such as in broadcasting services, a high-efficiency compression coding system is necessary. Therefore, we have developed a new SHV real-time hardware codec system maintains high picture quality by using eight 1080/60p (60 frames per second) encoding units and a video format converter with signal compensation processing. The new system uses about half the bit-rate and is about 50% smaller than the previous one.

CiNii Research

researchmap

Language：Japanese   Publisher：The Institute of Electronics, Information and Communication Engineers  

We have been developing a novel coding framework, "reconstructive video coding", that performs image size reduction and super-resolution to decrease coding degradation. In this paper, we propose a new super-resolution method that makes use of the local similarities of the image. The proposed method is used in the coding framework to improve the quality of the reconstructed image. The experiment shows the improvement of a subjective image quality, especially in texture areas and edge regions exploiting their similarities.

CiNii Research

researchmap

Language：Japanese   Publisher：The Institute of Electronics, Information and Communication Engineers  

CiNii Research

researchmap

Language：Japanese   Publisher：Forum on Information Technology  

CiNii Research

researchmap

Language：Japanese   Publisher：The Institute of Image Information and Television Engineers  

We propose a novel video coding method using wavelet super resolution (SR) for high compression ultrahigh video coding. The proposed method encodes reduced-size images and carries out SR reconstruction after decoding procedure. The SR reconstruction uses some parameters which are encoded as auxiliary information. The auxiliary information is determined by comparing a local-decoded SR image with the original image in encoder side.

CiNii Research

researchmap

Language：Japanese   Publisher：Forum on Information Technology  

CiNii Research

researchmap

Language：Japanese   Publisher：Forum on Information Technology  

CiNii Research

researchmap

Language：Japanese   Publisher：The Institute of Image Information and Television Engineers  

This paper proposes a system to personalize broadcasting TV programs using the Internet-delivered additional information. In this system the Internet-delivered content are provided in sync with the on-air broadcating and both broadcast and Internet content are re-synchronized and combined at the receiver before presentation. Functional requirements for the system and a prototype implementation of a personalized subtitle service are discussed.

CiNii Research

researchmap

Language：Japanese   Publisher：Forum on Information Technology  

CiNii Research

researchmap

Language：Japanese   Publisher：The Institute of Electronics, Information and Communication Engineers  

IP transmission has been widely deployed in broadcasting. Otherwise, it is difficult to transmit MPEG-2 Transport Stream to comply with MPEG-2 Systems requirements in which the "error of Program Clock Reference including delay jitter" should be suppressed to less than ±500ns. In this report, we illustrate reasons of the delay jitter and introduce a manner to fix this jitter by the Constant Delay Queuing (CDQ). Evaluating an experimental hub with the CDQ scheduler, it is demonstrated that the CDQ scheduler forwards prioritized frames within jitter less than 250ns and the experimental hub enables transmission of MPEG-2 Transport Stream over IP network within the MPEG-2 Systems requirement.

CiNii Research

researchmap

Language：Japanese   Publisher：The Institute of Electronics, Information and Communication Engineers  

We propose an intelligence circulation system which collects viewer's responses about TV program, analyzes those responses, and creates feedback services. The system consists of three parts as input-part, data analysis and reserve-part, and feedback service-part. We clarified the problems of each part, and implemented the system experimentally.

CiNii Research

researchmap

Language：Japanese   Publisher：Forum on Information Technology  

CiNii Research

researchmap

Language：Japanese   Publisher：日本ITU協会  

CiNii Research

researchmap

Language：Japanese   Publisher：The Institute of Electronics, Information and Communication Engineers  

A user-friendly information presentation system has been developed that enables visually impaired people to enjoy access to data broadcast and Internet content. The goals of this work were to develop a scheme for converting broadcast and communications content to a format readily accessible by the visually impaired, and to develop a scheme offering a range of output options that the user can freely choose from among to reflect the type and degree of disability. The report begins with a description of the system's configuration. Then we evaluate the system objectively by comparing the time it takes to search for actual data and Web content using the proposed tactile navigation versus conventional audio navigation, and subjectively based on user assessments of operability, effectiveness, content comprehension. The findings indicate that the tactile navigation approach has discernable learning-effect and content-comprehension-effect advantages.

CiNii Research

researchmap

Language：Japanese   Publisher：Information Processing Society of Japan (IPSJ)  

In recent years, due to the digitalisation of broadcast and the increase of broadcast channels, the number of news delivered to the user has multiplied more and more. Additionally, the user's preference has been diversifying rapidly. Therefore, authors have been constructing a news digest system extracting only necessary news topics from some news programs and laying out them in the order of their necessity. The previous works of creating digests have mainly dealt with estimating the user's preference accurately, and extracting the topics adapting to the user's preference correctly. However, if the users are provided only the topics adapting to their preferences, they might miss the important information such as hot topics and emergency news. Consequently, we proposed a method of creating news digests adapting to both the user's preference and the news importance. This paper focuses and evaluates the method of calculating news value by comparing the values which the system calculates to those which the executive editors of news programs judge. From the evaluation results, it is revealed that the accuracy of the method is more than 80%.

CiNii Research

researchmap

Language：Japanese   Publisher：The Institute of Electronics, Information and Communication Engineers  

AdapTV services have been proposed to present TV programs in various forms depending on viewer's context, such as circumstances, abilities and devices. This report describes a terminal-adaptive video trimming system aiming to improve visibility as a service model of the AdapTV. The proposed system classifies "long shot" and "medium shot", whose visibility expected to improve by trimming, according to color features. Then, the system trims players areas extracted from a video using luminance value with a simple image processing. The system can produce an easy-to-watch trimmed video with a low computational complexity.

CiNii Research

researchmap

Language：Japanese   Publisher：日本放送出版協会  

CiNii Research

researchmap

Language：Japanese  

CiNii Research

researchmap

Language：Japanese   Publisher：Forum on Information Technology  

CiNii Research

researchmap

Language：Japanese   Publisher：日本放送出版協会  

CiNii Research

researchmap

Language：Japanese  

CiNii Research

researchmap

Language：Japanese   Publisher：The Institute of Electronics, Information and Communication Engineers  

Recently, broadband internet have spread widely and video streaming service have come into wide use. For more enhancing of service, improving streaming quality, varieties of contents and channels are required. On the other hand, multi-channel streaming courses packets loss by their own burstiness. In this paper, we have proposed a stream shaping algorithm for multi-channel streaming, which is effective in eliminating the burstiness. It controls output timing of each stream and output timing of mutual streams at the same time. We have implemented this shaping algorithm by software and confirmed the effectiveness of this algorithm for reducing buffer fullness of network nodes.

CiNii Research

researchmap

Language：Japanese  

CiNii Research

researchmap

Language：Japanese   Publisher：Information Processing Society of Japan (IPSJ)  

The number of TV broadcasting channels has rapidly grown because of digitalization. This makes viewers difficult to select channels among them. A TV navigation system has thus been developed that enables a variety of viewers to enjoy TV programs in accordance with occasional mood and preferences with easy operations. The experimental TV navigation system developed provides combined facilities of recommending programs suited for viewers' daily life plan, presenting the program timetable personalized for their interests and retrieving program information in advance using names of persons and places. Evaluation tests have been accomplished on the experimental system with actual broadcasting contents.

CiNii Research

researchmap

Other Link： http://id.nii.ac.jp/1001/00041222/

Language：Japanese   Publisher：The Institute of Electronics, Information and Communication Engineers  

The number of TV broadcasting channels has rapidly grown because of digitalization. This makes viewers difficult to select channels among them. A TV navigation system has thus been developed that enables a variety of viewers to enjoy TV programs in accordance with occasional mood and preferences with easy operations. The experimental TV navigation system developed provides combined facilities of recommending programs suited for viewers' daily life plan, presenting the program timetable personalized for their interests and retrieving program information in advance using names of persons and places. Evaluation tests have been accomplished on the experimental system with actual broadcasting contents.

CiNii Research

researchmap

Language：Japanese  

CiNii Research

researchmap

Language：Japanese   Publisher：The Institute of Electronics, Information and Communication Engineers  

We have developed an experimental TV navigation system to assist viewers in selecting appropriate TV programs even when they are many digital broadcasting channels. This report describes several new results of an evaluation test for Remote Control Pointing Tools to find TV program easily. We used four typical pointing devices for the renewal system with on-screen EPG at HDTV resolution.

CiNii Research

researchmap

Language：Japanese   Publisher：The Institute of Electronics, Information and Communication Engineers  

The number of TV broadcasting channels has rapidly grown because of digitalization. This makes viewers difficult to select channels among them. A TV navigation system has thus been developed that enables a variety of viewers to enjoy TV programs in accordance with occasional mood and preferences with easy operations. The experimental TV navigation system developed provides combined facilities of recommending programs suited for viewers' daily life plan, presenting the program timetable personalized for their interests and retrieving program information in advance using names of persons and places. Evaluation tests have been accomplished on the experimental system with actual broadcasting contents.

CiNii Research

researchmap

Language：Japanese   Publisher：The Institute of Electronics, Information and Communication Engineers  

Broadband internet services have come into wide use recently. It also accelerates popularization of high qualityvideo streaming services. This paper describes an experimental system for video stream splicing and multiplexing at relayservers. MPEG-4 Studio Profile coding schemes are introduced. The video streaming system would provide more varietyof composition of TV programs and would enable to more personal or more local services.

CiNii Research

researchmap

Language：Japanese   Publisher：The Institute of Electronics, Information and Communication Engineers  

TV broadcasting over IP network has been studied as IP network has come into wide use. Network trouble like packet losses, which sometimes occurs in IP network, may lead to serious deterioration of pictures and audio quality of the transmitted TV programs. Considering the importance of broadcasting, what programs say must be transmitted as much as possible even if network trouble occurs. We study a system which transmits the most important content of programs at lower bit rate to cope with network trouble. We produce some alternatives of various programs, since the most important content of programs is different depending on their category. We evaluate them in terms of viewer's satisfaction. The result of assessment tests shows that the most preferable alternative changes depending on the extent of network trouble and also on the category of its original program.

CiNii Research

researchmap

Language：Japanese   Publisher：日本工業出版  

CiNii Research

researchmap

Language：Japanese   Publisher：日本工業出版  

CiNii Research

researchmap

Language：Japanese   Publisher：日本工業出版  

CiNii Research

researchmap

Language：Japanese   Publisher：日本工業出版  

CiNii Research

researchmap

Language：Japanese   Publisher：日本工業出版  

CiNii Research

researchmap

Language：English   Publisher：The Institute of Electronics, Information and Communication Engineers  

This paper describes a motion estimation method with eigenvalue algorithm matching local image feature. We introduce two estimation methods based on eigen value algorithm for improvement in accuracy. The first method is based on robust estimation. We apply this method to the block which includes multiple movements. The second method is applied to the block in which one typical gradient alone exists. In this method, we complement the eigenvector from the object with the same edge in the subjective block, and then motion estimation is calculated with the complementary eigenvector. Experimental results show that the proposed method improves the estimation accuracy.

CiNii Research

researchmap

Language：Japanese   Publisher：The Institute of Image Information and Television Engineers  

MPEG-2の後を受けて次世代の符号化規格として標準化が進んできたMPEG-4は, 通称バージョン1, バージョン2規格がすでに成立しており, さらに拡張が行われている.その中で, MPEG-4の特長であるオブジェクト符号化を放送局の映像制作に応用するとともに, これまでの規格がカバーしていなかった高画質映像を扱える符号化規格を作成するために, スタジオプロファイルの審議が続けられてきた.本稿では, MPEGでの審議を終了し, 正式な規格となるMPEG-4スタジオプロファイルについて, リクワイアメント, 仕様, 要素技術, 応用の各面について紹介する.

DOI： 10.3169/itej.55.822

CiNii Research

researchmap

Language：Japanese   Publisher：The Institute of Electronics, Information and Communication Engineers  

In order to improve the efficiency of video program production and to produce interactive programs, we propose a scene description language for video program production in broadcasting station. This paper describes advantages of program production using scene description language, requirements of program production and features of the proposed language. The proposed language is based on XML and defines elements for each function of program production. We have developed a software player for the proposed language and conducted program productions experiments.

CiNii Research

researchmap

Language：Japanese   Publisher：The Institute of Image Information and Television Engineers  

DOI： 10.3169/itej.55.511

CiNii Research

researchmap

Language：Japanese   Publisher：The Institute of Electronics, Information and Communication Engineers  

This paper proposes a lossless shape coding method. The coding scheme is based on the arithmetic coding for the wavelet descriptor. In this method, the wavelet descriptor is obtained by the wavelet transformation for the contour direction vectors along the shape contour. In order to encode the moving shape efficiently, we also define motion parameter detected by the matching method for the contour direction vectors. The simulation result shows that the proposed shape coding method is more efficient than that of MPEG-4.

CiNii Research

researchmap

Language：Japanese   Publisher：The Institute of Electronics, Information and Communication Engineers  

This paper describes novel object model for interoperability of contents on the distributed environment which many devices are connected on the network in TV program production. We also illustrate with the video player, which is a prototype model on the distributed processing environment, based on the proposed object model. In this system, we can playback the video contents on the network without reflecting the format, since the functional modules are automatically combined corresponding to the contents.

CiNii Research

researchmap

Language：Japanese   Publisher：The Institute of Electronics, Information and Communication Engineers  

CiNii Research

researchmap

Language：Japanese   Publisher：The Institute of Image Information and Television Engineers  

CiNii Research

researchmap

Language：Japanese   Publisher：The Institute of Image Information and Television Engineers  

CiNii Research

researchmap

Language：Japanese   Publisher：日本放送出版協会  

資料形態 : テキストデータ プレーンテキスト
コレクション : 国立国会図書館デジタルコレクション > デジタル化資料 > 雑誌

CiNii Research

researchmap

Language：Japanese   Publisher：日本放送出版協会  

CiNii Research

researchmap

Language：Japanese   Publisher：Information Processing Society of Japan (IPSJ)  

CiNii Research

researchmap

Language：Japanese  

researchmap

Language：Japanese   Publisher：The Institute of Electronics, Information and Communication Engineers  

This paper describes a motion estimation method with eigenvalue algorithm matching local image feature. The proposed method first formulates the optical flow using the block gradient method with eigenvalue algorithm, and then low reliability optical flow is classified according to the characteristics inside the block. We introduce two estimation methods based on eigenvalue algorithm for improvement in accuracy. The first method is based on robust estimation. We apply this method to the blocks which include multiple movements. The second method is applied to the blocks in which one typical gradient alone exists. In the method, we complement the eigenvector from the object with the same edge in the subjective block, and then motion estimation is carried out with the complementary eigenvector. Experimental results show that the proposed method improves the estimation accuracy.

CiNii Research

researchmap

Language：Japanese   Publisher：Information Processing Society of Japan (IPSJ)  

MPEG-4 version 1 (ISO/IEC 14496) has already been finalized as an international standard and version 2 is now under discussion. The visual part of version 1 features coding efficiency and error resilience functionality for very low bitrate video coding, and object-based video coding for arbitraily shaped video objects. The object-based coding functionality is useful for contents creation environments such as TV studios. In order to use MPEG-4 in the studio, the standard should be expanded on after version 2 to include high-quality video and effective handling properties. This paper reports requirements and current status of specifications of the MPEG-4 studio profile, and the results of preliminary experiments in the process of standardization.

CiNii Research

researchmap

Other Link： http://id.nii.ac.jp/1001/00041575/

Language：English  

Scopus

J-GLOBAL

researchmap

Language：Japanese   Publisher：The Institute of Image Information and Television Engineers  

CiNii Research

researchmap

Language：Japanese   Publisher：日本放送出版協会  

CiNii Research

researchmap

Language：Japanese   Publisher：日本放送出版協会  

CiNii Research

researchmap

Language：English   Publisher：The Institute of Image Information and Television Engineers  

CiNii Research

researchmap

Language：Japanese   Publisher：The Institute of Electronics, Information and Communication Engineers  

There has been many studies of improvements of the picture quality obtained by MPEG-2 video coding. It is possible to encode images efficiently by an appropriate allocation of the number of the bits to I, P, B picture types. In this paper, we first estimate the distortion from both rate-distortion properties for the different coded picture types and effect that the component of distortion in predicted images affects the coding efficiency of the corresponding images. Then we allocate the number of bits to each picture type to minimize the distortion. We examine this method by the MPEG-2 MP@HL algorithm at a bit-rate of 18Mbits/s using three sequences. Simulation results show that this method is superior to the TM5 method in SNR.

CiNii Research

researchmap

Language：Japanese   Publisher：The Institute of Electronics, Information and Communication Engineers  

CiNii Research

researchmap

Language：Japanese   Publisher：The Institute of Electronics, Information and Communication Engineers  

CiNii Research

researchmap

Language：Japanese   Publisher：日本放送出版協会  

CiNii Research

researchmap

Language：Japanese   Publisher：The Institute of Electronics, Information and Communication Engineers  

Estimating motions in image sequences is important in many areas of image processing, for example, in image segmentation. The conventional block matching generates unreliable motion fields in the sense of true motion in the scene, especially when using a small block size. In this paper we propose to perform motion estimation with matching of spatiotemporal blocks, which are packs of 2D spatial blocks taken from successive frames of an image sequence. Our simulations shows that motion estimation using spatiotemporal block matching is more robust than the 2D approaches and the obtained motion fields can represent the motions in the scene more accurately.

CiNii Research

researchmap

Language：Japanese   Publisher：Information Processing Society of Japan (IPSJ)  

Object-based picture coding applies the required segmentation method to classify motion pictures into objects. The motion estimation technique provides essential data in the segmentation method, are it is expected to provide accurate motion estimation. The gradient based motion estimation integrating the eigenvalue algorithm has the capacity to estimate large movements which are beyond the ability of the least square algorithm to measure accurately. Eigenvalues that are obtained by the use of this algorithm can represent reliability indices of the estimated motion vector, and the reliability indices can be applied to improve the accuracy of motion estimation. This paper describes the characteristics of gradient based motion estimation with the eigenvalue algorithm in comparison with the least square algorithm The from this characteristic of the eigenvalue algorithm, we shall examine motion estimation with optimized block size using reliability indices.

CiNii Research

researchmap

Language：Japanese   Publisher：The Institute of Electronics, Information and Communication Engineers  

In this paper, we examine the motion estimation using gradient method, and propose a technique to improve the estimation accuracy. We first propose a reliability of estimated motion, and conform the properness of the proposed reliability through experiments. Then, we classify motion pictures using the reliability into regions, and characterize each of the classification regions. Finally, we try to improve the accuracy of the motion estimation by applying a method which fits the characteristic of the classified region. We have verified the effectiveness of the proposed technique by simulation.

CiNii Research

researchmap

Language：Japanese   Publisher：The Institute of Electronics, Information and Communication Engineers  

In this paper, we propose an image segmentation method which is used for a pre-processing of the object-based picture coding. The segmentation method by K-means algorithm, which is a widely used clustering method, sometimes segments an image into too many small regions and also splits the image at a position that does not correspond to the original boundary of an object in the image. To deal with these problems, we apply a three step region modifying scheme to the clustering result. We can obtain satisfactory results using this method.

CiNii Research

researchmap

Language：Japanese   Publisher：一般社団法人映像情報メディア学会  

CiNii Research

researchmap

Language：Japanese   Publisher：The Institute of Electronics, Information and Communication Engineers  

In this paper, we investigate the performance of a disparity estimation method called hierarchical block cross-correlation in an image coding context. We first propose a new, coding error based evaluation criterion. Next, using this new criterion, we compare the hierarchical block cross-correlation method with two other popular ones, full searching block matching and hierchical block matching. The simulation results show that, compared to other methods, the hierarchical block cross-correlation gives comparable or smaller coding errors while resulting in smoother disparity vectors.

CiNii Research

researchmap

Language：Japanese   Publisher：The Institute of Electronics, Information and Communication Engineers  

CiNii Research

researchmap

Language：Japanese   Publisher：The Institute of Image Information and Television Engineers  

This paper proposes a new coding technique, termed "Region Support DCT", for an arbitrary shape coding. The technique is based on the conventional two-dimensional DCT, and, therefore, the decoding process is achieved by the conventional one as it is. The technique outperforms other techniques for arbitrary shape coding, such as SA-DCT.

CiNii Research

researchmap

Language：Japanese   Publisher：The Institute of Image Information and Television Engineers  

CiNii Research

researchmap

Language：Japanese   Publisher：一般社団法人映像情報メディア学会  

CiNii Research

researchmap

Language：Japanese   Publisher：日本放送出版協会  

CiNii Research

researchmap

Language：Japanese   Publisher：The Institute of Image Information and Television Engineers  

Bi-directional prediction scheme has been adopted by MPEG as a standardized coding technique. Hitherto, the one directional prediction scheme, which employs temporally pre-frame alone, has been widely used. Recently, the bi-directional prediction scheme, which uses both temporally pre and post frames, is also utilized. It is generally known that the bi-directional prediction is effective with respect to the coding efficiency. However, the effectiveness has yet to be examined through subjective picture quality assessment. we conducted a subjective assessment test of the decoded picture quality using HDTV sequences. The results show that the bi-directional prediction scheme works Well on most of the sequences with respect to picture quality improvement, but causes flickers to some sequences.

CiNii Research

researchmap

Language：Japanese   Publisher：一般社団法人映像情報メディア学会  

CiNii Research

researchmap

Language：Japanese   Publisher：日本放送出版協会  

CiNii Research

researchmap

Language：Japanese   Publisher：一般社団法人映像情報メディア学会  

CiNii Research

researchmap

Language：Japanese   Publisher：一般社団法人映像情報メディア学会  

CiNii Research

researchmap

Language：Japanese   Publisher：一般社団法人映像情報メディア学会  

CiNii Research

researchmap

Language：Japanese   Publisher：The Institute of Image Information and Television Engineers  

This paper discusses the characteristics of Hi-Vision codecs using a hybrid vector quantization connected in tandem.

CiNii Research

researchmap

Language：English   Publisher：一般社団法人映像情報メディア学会  

CiNii Research

researchmap

Language：Japanese   Presentation type：Oral presentation (general)  

researchmap

Language：Japanese   Presentation type：Oral presentation (general)  

researchmap

Language：Japanese  

researchmap

Language：Japanese  

researchmap

Language：Japanese  

researchmap

Language：Japanese  

researchmap

Language：Japanese  

researchmap

Language：Japanese  

researchmap

Language：Japanese  

researchmap

Language：Japanese  

researchmap

Language：Japanese  

researchmap

Language：Japanese  

researchmap

Language：Japanese  

researchmap

Language：Japanese  

researchmap

Language：Japanese  

researchmap

Language：Japanese  

researchmap

Language：Japanese  

researchmap

Language：Japanese  

researchmap

Language：Japanese  

researchmap

Language：Japanese  

The Super Hi-Vision (SHV) is an ultra high-definition video system with 4,000 scanning lines. Its video format is 7,680 x 4,320 pixels and the frame rate is 60 Hz with progressive scanning. It has been designed to give viewers a strong sense of reality. To make the system suitable for practical use, such as in broadcasting services, a high-efficiency compression coding system is necessary. Therefore, we have developed a new SHV real-time hardware codec system maintains high picture quality by using eight 1080/60p (60 frames per second) encoding units and a video format converter with signal compensation processing. The new system uses about half the bit-rate and is about 50% smaller than the previous one.

researchmap

Language：Japanese  

We have been developing a novel coding framework, "reconstructive video coding", that performs image size reduction and super-resolution to decrease coding degradation. In this paper, we propose a new super-resolution method that makes use of the local similarities of the image. The proposed method is used in the coding framework to improve the quality of the reconstructed image. The experiment shows the improvement of a subjective image quality, especially in texture areas and edge regions exploiting their similarities.

researchmap

Language：Japanese  

researchmap

Language：Japanese  

researchmap

Language：Japanese  

researchmap

Language：Japanese  

researchmap

Language：Japanese  

This paper proposes a novel iterative super-resolution method that generates high-frequency information for an enlarged image. The block-wise high-frequency information is obtained through a simulated process of down-scaling and up-scaling of a corresponding larger homothetic block. The experimental results show the effectiveness of the proposed

researchmap

Language：Japanese  

We propose a novel video coding method using wavelet super resolution (SR) for high compression ultrahigh video coding. The proposed method encodes reduced-size images and carries out SR reconstruction after decoding procedure. The SR reconstruction uses some parameters which are encoded as auxiliary information. The auxiliary information is determined by comparing a local-decoded SR image with the original image in encoder side.

researchmap

Language：Japanese  

We propose a novel intra prediction method. It applies a discrete cosine transform and an inversely transform to an image region repeatedly to obtains better intra prediction pixel values.

researchmap

Language：Japanese  

researchmap

Language：Japanese  

IP transmission has been widely deployed in broadcasting. Otherwise, it is difficult to transmit MPEG-2 Transport Stream to comply with MPEG-2 Systems requirements in which the "error of Program Clock Reference including delay jitter" should be suppressed to less than ±500ns. In this report, we illustrate reasons of the delay jitter and introduce a manner to fix this jitter by the Constant Delay Queuing (CDQ). Evaluating an experimental hub with the CDQ scheduler, it isdemonstrated that the CDQ scheduler forwards prioritized frames within jitter less than 250ns and the experimental hub enables transmission of MPEG-2 Transport Stream over IP network within the MPEG-2 Systems requirement.

researchmap

Language：Japanese  

IP transmission has been widely deployed in broadcasting. Otherwise, it is difficult to transmit MPEG-2 Transport Stream to comply with MPEG-2 Systems requirements in which the "error of Program Clock Reference including delay jitter" should be suppressed to less than ±500ns. In this report, we illustrate reasons of the delay jitter and introduce a manner to fix this jitter by the Constant Delay Queuing (CDQ). Evaluating an experimental hub with the CDQ scheduler, it is demonstrated that the CDQ scheduler forwards prioritized frames within jitter less than 250ns and the experimental hub enables transmission of MPEG-2 Transport Stream over IP network within the MPEG-2 Systems requirement.

researchmap

Language：Japanese  

We propose an intelligence circulation system which collects viewer's responses about TV program, analyzes those responses, and creates feedback services. The system consists of three parts as input-part, data analysis and reserve-part, and feedback service-part. We clarified the problems of each part, and implemented the system experimentally.

researchmap

Language：Japanese  

researchmap

Language：Japanese  

A new graphical user interface is proposed. With proposal interface, user can select information from a number of options using a remote controller with a trackball viewing on TV display. Information options are allayed in lines in 3-dimensional space and displaying size and position of each option are changed according to user's control.

researchmap

Language：Japanese  

This paper proposes a system to personalize broadcasting TV programs using the Internet-delivered additional information. In this system the Internet-delivered content are provided in sync with the on-air broadcating and both broadcast and Internet content are re-synchronized and combined at the receiver before presentation. Functional requirements for the system and a prototype implementation of a personalized subtitle service are discussed.

researchmap

Language：Japanese  

A user-friendly information presentation system has been developed that enables visually impaired people to enjoy access to data broadcast and Internet content. The goals of this work were to develop a scheme for converting broadcast and communications content to a format readily accessible by the visually impaired, and to develop a scheme offering a range of output options that the user can freely choose from among to reflect the type and degree of disability. The report begins with a description of the system's configuration. Then we evaluate the system objectively by comparing the time it takes to search for actual data and Web content using the proposed tactile navigation versus conventional audio navigation, and subjectively based on user assessments of operability, effectiveness, content comprehension. The findings indicate that the tactile navigation approach has discernable learning-effect and content-comprehension-effect advantages.

researchmap

Language：Japanese  

In recent years, due to the digitalisation of broadcast and the increase of broadcast channels, the number of news delivered to the user has multiplied more and more. Additionally, the user's preference has been diversifying rapidly. Therefore, authors have been constructing a news digest system extracting only necessary news topics from some news programs and laying out them in the order of their necessity. The previous works of creating digests have mainly dealt with estimating the user's preference accurately, and extracting the topics adapting to the user's preference correctly. However, if the users are provided only the topics adapting to their preferences, they might miss the important information such as hot topics and emergency news. Consequently, we proposed a method of creating news digests adapting to both the user's preference and the news importance. This paper focuses and evaluates the method of calculating news value by comparing the values which the system calculates to those which the executive editors of news programs judge. From the evaluation results, it is revealed that the accuracy of the method is more than 80%.

researchmap

Language：Japanese  

In recent years, due to the digitalisation of broadcast and the increase of broadcast channels, the number of news delivered to the user has multiplied more and more. Additionally, the user's preference has been diversifying rapidly. Therefore, authors have been constructing a news digest system extracting only necessary news topics from some news programs and laying out them in the order of their necessity. The previous works of creating digests have mainly dealt with estimating the user's preference accurately, and extracting the topics adapting to the user's preference correctly. However, if the users are provided only the topics adapting to their preferences, they might miss the important information such as hot topics and emergency news. Consequently, we proposed a method of creating news digests adapting to both the user's preference and the news importance. This paper focuses and evaluates the method of calculating news value by comparing the values which the system calculates to those which the executive editors of news programs judge. From the evaluation results, it is revealed that the accuracy of the method is more than 80%.

researchmap

Language：Japanese  

AdapTV services have been proposed to present TV programs in various forms depending on viewer's context, such as circumstances, abilities and devices. This report describes a terminal-adaptive video trimming system aiming to improve visibility as a service model of the AdapTV. The proposed system classifies "long shot" and "medium shot", whose visibility expected to improve by trimming, according to color features. Then, the system trims players areas extracted from a video using luminance value with a simple image processing. The system can produce an easy-to-watch trimmed video with a low computational complexity.

researchmap

Language：Japanese  

researchmap

Language：Japanese  

researchmap

Language：Japanese  

researchmap

Language：Japanese  

Recently, broadband internet have spread widely and video streaming service have come into wide use. For more enhancing of service, improving streaming quality, varieties of contents and channels are required. On the other hand, multi-channel streaming courses packets loss by their own burstiness. In this paper, we have proposed a stream shaping algorithm for multi-channel streaming, which is effective in eliminating the burstiness. It controls output timing of each stream and output timing of mutual streams at the same time. We have implemented this shaping algorithm by software and confirmed the effectiveness of this algorithm for reducing buffer fullness of network nodes.

researchmap

Language：Japanese  

researchmap

Language：Japanese  

Streaming contents of TV programs over IP network may not be received just as broadcasting corporation intended. It is a result of network trouble like packet losses. This is one of the problems to be solved to realize TV broadcasting over IP network. This paper describes an experimental system which keeps broadcasting the information of programs even if network trouble occurs.

researchmap

Language：Japanese  

The spread of broadband IP networks is making high-quality video distribution service a reality. However, some technical problems should be solved to deliver programs nationwide with quality and reliability equivalent to that of current broadcast services using radio wave. This study illustrates high level requirements of realizing technology of high quality and dependable IP broadcasting service.

researchmap

Language：Japanese  

The number of TV broadcasting channels has rapidly grown because of digitalization. This makes viewers difficult to select channels among them. A TV navigation system has thus been developed that enables a variety of viewers to enjoy TV programs in accordance with occasional mood and preferences with easy operations. The experimental TV navigation system developed provides combined facilities of recommending programs suited for viewers' daily life plan, presenting the program timetable personalized for their interests and retrieving program information in advance using names of persons and places. Evaluation tests have been accomplished on the experimental system with actual broadcasting contents.

researchmap

Language：Japanese  

The number of TV broadcasting channels has rapidly grown because of digitalization. This makes viewers difficult to select channels among them. A TV navigation system has thus been developed that enables a variety of viewers to enjoy TV programs in accordance with occasional mood and preferences with easy operations. The experimental TV navigation system developed provides combined facilities of recommending programs suited for viewers' daily life plan, presenting the program timetable personalized for their interests and retrieving program information in advance using names of persons and places. Evaluation tests have been accomplished on the experimental system with actual broadcasting contents.

researchmap

Language：Japanese  

The number of TV broadcasting channels has rapidly grown because of digitalization. This makes viewers difficult to select channels among them. A TV navigation system has thus been developed that enables a variety of viewers to enjoy TV programs in accordance with occasional mood and preferences with easy operations. The experimental TV navigation system developed provides combined facilities of recommending programs suited for viewers' daily life plan, presenting the program timetable personalized for their interests and retrieving program information in advance using names of persons and places. Evaluation tests have been accomplished on the experimental system with actual broadcasting contents.

researchmap

Language：Japanese  

researchmap

Language：Japanese  

The number of TV broadcasting channels has rapidly grown because of digitalization. This makes viewers difficult to select channels among them. A TV navigation system has thus been developed that enables a variety of viewers to enjoy TV programs in accordance with occasional mood and preferences with easy operations. The experimental TV navigation system developed provides combined facilities of recommending programs suited for viewers' daily life plan, presenting the program timetable personalized for their interests and retrieving program information in advance using names of persons and places. Evaluation tests have been accomplished on the experimental system with actual broadcasting contents.

researchmap

Language：Japanese  

We have developed an experimental TV navigation system to assist viewers in selecting appropriate TV programs even when they are many digital broadcasting channels. This report describes several new results of an evaluation test for Remote Control Pointing Tools to find TV program easily. We used four typical pointing devices for the renewal system with on-screen EPG at HDTV resolution.

researchmap

Language：Japanese  

Broadband internet services have come into wide use recently. It also accelerates popularization of high quality video streaming services. This paper describes an experimental system for video stream splicing and multiplexing at relay servers. MPEG-4 Studio Profile coding schemes are introduced. The video streaming system would provide more variety of composition of TV programs and would enable to more personal or more local services.

researchmap

Language：Japanese  

TV broadcasting over IP network has been studied as IP network has come into wide use. Network trouble like packet losses, which sometimes occurs in IP network, may lead to serious deterioration of pictures and audio quality of the transmitted TV programs. Considering the importance of broadcasting, what programs say must be transmitted as much as possible even if network trouble occurs. We study a system which transmits the most important content of programs at lower bit rate to cope with network trouble. We produce some alternatives of various programs, since the most important content of programs is different depending on their category. We evaluate them in terms of viewer's satisfaction. The result of assessment tests shows that the most preferable alternative changes depending on the extent of network trouble and also on the category of its original program.

researchmap

Language：Japanese  

TV broadcasting over IP network has been studied as IP network has come into wide use. Network trouble like packet losses, which sometimes occurs in IP network, may lead to serious deterioration of pictures and audio quality of the transmitted TV programs. Considering the importance of broadcasting, what programs say must be transmitted as much as possible even if network trouble occurs. We study a system which transmits the most important content of programs at lower bit rate to cope with network trouble. We produce some alternatives of various programs, since the most important content of programs is different depending on their category. We evaluate them in terms of viewer's satisfaction. The result of assessment tests shows that the most preferable alternative changes depending on the extent of network trouble and also on the category of its original program.

researchmap

Language：Japanese  

Broadband internet services have come into wide use recently. It also accelerates popularization of high qualityvideo streaming services. This paper describes an experimental system for video stream splicing and multiplexing at relayservers. MPEG-4 Studio Profile coding schemes are introduced. The video streaming system would provide more varietyof composition of TV programs and would enable to more personal or more local services.

researchmap

Language：Japanese  

Broadband IP network is gradually making feasible video distribution services. Studying how to transmit TV programs if 1P network trouble like packet losses occurs is important. We propose a system which controls an amount of adding redundancy depending on the extent of network trouble to transmit at least what TV programs say. In this paper we make an estimate of an amount of adding redundancy to ensure target quality at a receiver.

researchmap

BS and 110E CS digital broadcasting services began in December 2000 and March 2002 respectively, and terrestrial digital broadcasting services will begin in December 2003. At that time we will need an advanced and interactive receiving navigation system to quickly find out and to watch desired programs or services for various TV viewers, including elderly and handicapped people. Here are some results of a study on a pointing method to select items on the TV screen, which provides an interactive operation scheme with universal human interface to easily select the name of programs and services.

researchmap

Language：Japanese  

A new TV navigation system to support viewing TV programs was developed and subjective evaluation tests were carried out. It recommends optimum programs depending on viewing habit and viewers profile besides viewers' easy program selection.

researchmap

researchmap

researchmap

Language：Japanese  

In order to improve the efficiency of video program production and to produce interactive programs, we propose a scene description language for video program production in broadcasting station. This paper describes advantages of program production using scene description language, requirements of program production and features of the proposed language. The proposed language is based on XML and defines elements for each function of program production. We have developed a software player for the proposed language and conducted program productions experiments.

researchmap

Language：Japanese  

This paper proposes a lossless shape coding method. The coding scheme is based on the arithmetic coding for the wavelet descriptor. In this method, the wavelet descriptor is obtained by the wavelet transformation for the contour direction vectors along the shape contour. In order to encode the moving shape efficiently, we also define motion parameter detected by the matching method for the contour direction vectors. The simulation result shows that the proposed shape coding method is more efficient than that of MPEG-4.

researchmap

Language：Japanese  

This paper proposes a lossless shape coding method. The coding scheme is based on the arithmetic coding for the wavelet descriptor. In this method, the wavelet descriptor is obtained by the wavelet transformation for the contour direction vectors along the shape contour. In order to encode the moving shape efficiently, we also define motion parameter detected by the matching method for the contour direction vectors. The simulation result shows that the proposed shape coding method is more efficient than that of MPEG-4.

researchmap

Language：Japanese  

This paper describes novel object model for interoperability of contents on the distributed environment which many devices are connected on the network in TV program production. We also illustrate with the video player, which is a prototype model on the distributed processing environment, based on the proposed object model. In this system, we can playback the video contents on the network without reflecting the format, since the functional modules are automatically combined corresponding to the contents.

researchmap

Language：Japanese  

researchmap

Language：Japanese  

This paper describes subjective evaluation results of low bit-rate video coding. The evaluation tests are based on the terms which are derived from the SD (Semantic Difference) and the factor analysis. The obtained data will be helpful to the design of MPEG-4 video encoding system in the digital terrestrial television broadcasting.

researchmap

Language：Japanese  

Although many investigations have been made on various aspects of motion-compensated frame differences, most of them assume that motion-compensated frame differences are homogeneous, However, it is observed that the variances of motion-compensated frame diffrernces at block boundaries tend to be larger than those at block centers. In this paper we present a theoretical interpretation of the space-dependent characteristics of motion-compensated frame differences based on a statistic motion distribution model. The model is then applied to analysis of prediction efficiency of overlapped block motion compensation. The analytical results are justified by empirical experiments with typical image sequences.

researchmap

Language：Japanese  

researchmap

Language：Japanese  

researchmap

Language：Japanese  

This paper proposes an inter frame lossless shape image coding method based on the contour direction vector expression. Motion vectors are detected by the matching method for the contour direction vectors. After wavelet transform, the contour direction vectors are encoded by the arithmetic coding. The simulation result shows that the proposed method is more efficient than that of MPEG-4.

researchmap

Language：Japanese  

This paper describes a motion estimation method with eigenvalue algorithm matching local image feature. The proposed method first formulates the optical flow using the block gradient method with eigenvalue algorithm, and then low reliability optical flow is classified according to the characteristics inside the block. We introduce two estimation methods based on eigenvalue algorithm for improvement in accuracy. The first method is based on robust estimation. We apply this method to the blocks which include multiple movements. The second method is applied to the blocks in which one typical gradient alone exists. In the method, we complement the eigenvector from the object with the same edge in the subjective block, and then motion estimation is carried out with the complementary eigenvector. Experimental results show that the proposed method improves the estimation accuracy.

researchmap

Language：Japanese  

This paper describes a motion estimation method with eigenvalue algorithm matching local image feature. The proposed method first formulates the optical flow using the block gradient method with eigenvalue algorithm, and then low reliability optical flow is classified according to the characteristics inside the block. We introduce two estimation methods based on eigenvalue algorithm for improvement in accuracy. The first method is based on robust estimation. We apply this method to the blocks which include multiple movements. The second method is applied to the blocks in which one typical gradient alone exists. In the method, we complement the eigenvector from the object with the same edge in the subjective block, and then motion estimation is carried out with the complementary eigenvector. Experimental results show that the proposed method improves the estimation accuracy.

researchmap

Language：Japanese  

This paper describes a motion estimation method with eigenvalue algorithm matching local image feature. The proposed method first formulates the optical flow using the block gradient method with eigenvalue algorithm, and then low reliability optical flow is classified according to the characteristics inside the block. We introduce two estimation methods based on eigenvalue algorithm for improvement in accuracy. The first method is based on robust estimation. We apply this method to the blocks which include multiple movements. The second method is applied to the blocks in which one typical gradient alone exists. In the method, we complement the eigenvector from the object with the same edge in the subjective block, and then motion estimation is carried out with the complementary eigenvector. Experimental results show that the proposed method improves the estimation accuracy.

researchmap

Language：Japanese  

MPEG-4 version 1 (ISO/IEC 14496) has already been finalized as an international standard and version 2 is now under discussion. The visual part of version 1 features coding efficiency and error resilience functionality for very low bitrate video coding, and object-based video coding for arbitraily shaped video objects. The object-based coding functionality is useful for contents creation environments such as TV studios. In order to use MPEG-4 in the studio, the standard should be expanded on after version 2 to include high-quality video and effective handling properties. This paper reports requirements and current status of specifications of the MPEG-4 studio profile, and the results of preliminary experiments in the process of standardization.

researchmap

Language：English  

Information about camera operating such as zoom, focus, pan, tilt and tracking is useful not only for efficient video coding, but also for content-based video representation. A camera operating parameter measurement system designed specifically for these applications is therefore developed.This system, implemented in real time and synchronized with the video signal, measures the precise camera operating parameters.We calibrated the camera lens using a camera model that accounts for radial lens distortion.The system is then applied to infer image motion from pan and tilt operating parameters.The experimental results show that the inferred motion coincides with the actual motion very well, with an error of less than 0.5 pixel even for large motion up to 80 pixels.

researchmap

Language：Japanese  

This paper describes a method of computational complexity reduction for the gradient motion estimation with eigenvalue algorithm. The proposed method adopts the algorithm based on perturbation theorem. The experimental results have proved that the proposed method can reduce the computational complexity and estimate movement with the same accuracy as the conventional algorithm.

researchmap

Language：Japanese  

This paper describes a motion estimation method with eigenvalue algorithm matching local image feature. The proposed method first formulates the optical flow using the block gradient method with eigenvalue algorithm, and then low reliability optical flow is classified according to the characteristics inside the block. We introduce two estimation methods based on eigenvalue algorithm for improvement in accuracy. The first method is based on robust estimation. We apply this method to the blocks which include multiple movements. The second method is applied to the blocks in which one typical gradient alone exists. In the method, we complement the eigenvector from the object with the same edge in the subjective block, and then motion estimation is carried out with the complementary eigenvector. Experimental results show that the proposed method improves the estimation accuracy.

researchmap

Language：Japanese  

researchmap

Language：English  

Information about camera operating such as zoom, focus, pan, tilt and tracking is useful not only for efficient video coding, but also for content-based video representation. A camera operating parameter measurement system designed specifically for these applications is therefore developed. This system, implemented in real time and synchronized with the video signal, measures the precise camera operating parameters. We calibrated the camera lens using a camera model that accounts for radial lens distortion. The system is then applied to infer image motion from pan and tilt operating parameters. The experimental results show that the inferred motion coincides with the actual motion very well, with an error of less than 0.5 pixel even for large motion up to 80 pixels.

researchmap

Language：Japanese  

researchmap

Language：Japanese  

researchmap

Language：Japanese  

There has been many studies of improvements of the picture quality obtained by MPEG-2 video coding. It is possible to encode images efficiently by an appropriate allocation of the number of the bits to I, P, B picture types. In this paper, we first estimate the distortion from both rate-distortion properties for the different coded picture types and effect that the component of distortion in predicted images affects the coding efficiency of the corresponding images. Then we allocate the number of bits to each picture type to minimize the distortion. We examine this method by the MPEG-2 MP@HL algorithm at a bit-rate of 18Mbits/s using three sequences. Simulation results show that this method is superior to the TM5 method in SNR.

researchmap

Language：Japanese  

There has been many studies of improvements of the picture quality obtained by MPEG-2 video coding. It is possible to encode images efficiently by an appropriate allocation of the number of the bits to I, P, B picture types. In this paper, we first estimate the distortion from both rate-distortion properties for the different coded picture types and effect that the component of distortion in predicted images affects the coding efficiency of the corresponding images. Then we allocate the number of bits to each picture type to minimize the distortion. We examine this method by the MPEG-2 MP@HL algorithm at a bit-rate of 18Mbits/s using three sequences. Simulation results show that this method is superior to the TM5 method in SNR.

researchmap

Language：Japanese  

researchmap

Language：English  

researchmap

Language：Japanese  

researchmap

Language：Japanese  

researchmap

Language：Japanese  

Estimating motions in image sequences is important in many areas of image processing, for example, in image segmentation. The conventional block matching generates unreliable motion fields in the sense of true motion in the scene, especially when using a small block size. In this paper we propose to perform motion estimation with matching of spatiotemporal blocks, which are packs of 2D spatial blocks taken from successive frames of an image sequence. Our simulations shows that motion estimation using spatiotemporal block matching is more robust than the 2D approaches and the obtained motion fields can represent the motions in the scene more accurately.

researchmap

Language：Japanese  

Object-based picture coding applies the required segmentation method to classify motion pictures into objects. The motion estimation technique provides essential data in the segmentation method, are it is expected to provide accurate motion estimation. The gradient based motion estimation integrating the eigenvalue algorithm has the capacity to estimate large movements which are beyond the ability of the least square algorithm to measure accurately. Eigenvalues that are obtained by the use of this algorithm can represent reliability indices of the estimated motion vector, and the reliability indices can be applied to improve the accuracy of motion estimation. This paper describes the characteristics of gradient based motion estimation with the eigenvalue algorithm in comparison with the least square algorithm The from this characteristic of the eigenvalue algorithm, we shall examine motion estimation with optimized block size using reliability indices.

researchmap

Language：Japanese  

In this paper, we examine the motion estimation using gradient method, and propose a technique to improve the estimation accuracy. We first propose a reliability of estimated motion, and conform the properness of the proposed reliability through experiments. Then, we classify motion pictures using the reliability into regions, and characterize each of the classification regions. Finally, we try to improve the accuracy of the motion estimation by applying a method which fits the characteristic of the classified region. We have verified the effectiveness of the proposed technique by simulation.

researchmap

Language：Japanese  

In this paper, we propose an image segmentation method which is used for a pre-processing of the object-based picture coding. The segmentation method by K-means algorithm, which is a widely used clustering method, sometimes segments an image into too many small regions and also splits the image at a position that does not correspond to the original boundary of an object in the image. To deal with these problems, we apply a three step region modifying scheme to the clustering result. We can obtain satisfactory results using this method.

researchmap

Language：Japanese  

researchmap

Language：Japanese  

In this paper, we investigate the performance of a disparity estimation method called hierarchical block cross-correlation in an image coding context. We first propose a new, coding error based evaluation criterion. Next, using this new criterion, we compare the hierarchical block cross-correlation method with two other popular ones, full searching block matching and hierchical block matching. The simulation results show that, compared to other methods, the hierarchical block cross-correlation gives comparable or smaller coding errors while resulting in smoother disparity vectors.

researchmap

Language：Japanese  

動画像処理の分野における領域分割や符号化等では動画像の正確な動き推定を行うことが重要である。動き推定の手法の一つである勾配法はマッチング法に比べ推定精度に制限がないという特徴を持つ。しかし、小さな動きに対しては誤差の小さい推定が可能であるが、大きな動きに対しては推定誤差が大きいことが欠点とされている。本報告では、勾配法による動き推定として、固有ベクトルを用いた計算手法の検討を行った。

researchmap

Language：Japanese  

We previously proposed RS(Region Support)-DCT, a novel coding tool for texture within an arbitrarily shaped region. In this paper, we study a derivation process of a RS-DCT coefficient set. The result shows that if a conventional type VLC method is assumed, an introduction of a frequency weighting into the coefficient derivation works well.

researchmap

Language：Japanese  

This paper proposes a new coding technique, termed "Region Support DCT", for an arbitrary shape coding. The technique is based on the conventional two-dimensional DCT, and, therefore, the decoding process is achieved by the conventional one as it is. The technique outperforms other techniques for arbitrary shape coding, such as SA-DCT.

researchmap

Language：Japanese  

researchmap

Language：Japanese  

This paper proposes a new coding technique, termed "Region Support DCT", for an arbitrary shape coding. The technique is based on the conventional two-dimensional DCT, and, therefore, the decoding process is achieved by the conventional one as it is. The technique outperforms other techniques for arbitrary shape coding, such as SA-DCT.

researchmap

Language：Japanese  

Bi-directional prediction scheme has been adopted by MPEG as a standardized coding technique.Hitherto,the one directional prediction scheme,which employs temporally pre-frame alone,has been widely used.Recently,the bi-directional prediction scheme, which uses both temporally pre and post frames,is also utilized.It is generally known that the bi-directional prediction is effective with respect to the coding efficiency.However,the effectiveness has yet to be examined through subjective picture quality assessment.We conducted a subjective assessment test of the decoded picture quality using HDTV sequences.The results show that the bidirectional prediction,scheme works well on most of the sequences with respect to picture quality improvement,but causes flickers to some sequences.

researchmap

Language：Japanese  

This paper describes a theoretical model of the power spectrum density(PSD)for the field-merged frame signal,and compares the efficiency between the field-based coding,and frame-based coding that uses the field-merged signal.Recently,for the coding of interlace-scanned television signals,the field-merged signal is also used as a input signal to spatial coding process.However,the nature of PSD,which is an important basis for a high efficiency coding,is not clarified enough for the field-merged signal.For instance,there is no assurrance of the optimality of the conventional techniques that are applied to the field-merged signal.The author has introduced a theoretical PSD model for the field.merged signal,and has clarified its properties in relation with the motion of the image.Using the model,a performance of frame based-coding has been calculated.

researchmap

Language：Japanese  

Bi-directional prediction scheme has been adopted by MPEG as a standardized coding technique. Hitherto, the one directional prediction scheme, which employs temporally pre-frame alone, has been widely used. Recently, the bi-directional prediction scheme, which uses both temporally pre and post frames, is also utilized. It is generally known that the bi-directional prediction is effective with respect to the coding efficiency. However, the effectiveness has yet to be examined through subjective picture quality assessment. we conducted a subjective assessment test of the decoded picture quality using HDTV sequences. The results show that the bi-directional prediction scheme works Well on most of the sequences with respect to picture quality improvement, but causes flickers to some sequences.

researchmap

Language：Japanese  

researchmap

Language：Japanese  

researchmap

Language：Japanese  

researchmap

Language：Japanese  

researchmap

Language：Japanese  

researchmap

Language：Japanese  

A study on a coding scheme which is appropriate for motion compensation prediction (MC) error signals is presented. MC error signals have very small correlation between adjacent pixels, especially in the vertical direction. Moreover, the correlation varies to a large extent depending on scene, local feature and type of movement. Therefore, with respect to coding gain and decoded picture quality, an adaptive coding scheme where orthogonal transform or subband partition is applied only for cases having high correlation is likely efficient for MC error signals.

researchmap

Language：Japanese  

Studies have been conducted for the development of an HDTV coding and transmission system for contribution purposes. This paper describes the results of investigations into loop filters used in the coding systems. HDTV signals contain a large amount of noise due to their wide bandwidth. This noise leads to poor coding efficiency. A pre-processing of coding input signals, however, is not preferable for high-fidelity transmission. Hence, the introduction of a loop-filter is examined, which reduces the noise in the prediction loop and increases the prediction accuracy. An investigation into a suitable filter characteristics is also conducted. The results show that the loop-filter is very effective for contribution-purpose coding systems, especially for ones using one pel accuracy motion compensation.

researchmap

Language：Japanese  

researchmap

Language：Japanese  

researchmap

Language：Japanese  

researchmap

researchmap

Language：Japanese  

An HDTV codec based on DCT plus motion-compensated prediction at a bitrate of 135 Mb/s is now being developed. The hardware prototype will be completed soon. This paper describes the specification and the configuration of the codec.

researchmap

Language：Japanese  

researchmap

Language：Japanese  

This paper discusses the characteristics of Hi-Vision codecs using a hybrid vector quantization connected in tandem.

researchmap

Language：Japanese  

researchmap

Language：Japanese  

researchmap

Language：Japanese  

researchmap

Language：Japanese  

researchmap

Applicant：日本放送協会

Application no：特願2013-214013  Date applied：2013.10

Announcement no：特開2015-076857  Date announced：2015.4

J-GLOBAL

researchmap

Applicant：日本放送協会

Application no：特願2013-214013  Date applied：2013.10

Announcement no：特開2015-076857  Date announced：2015.4

Patent/Registration no：特許第6212348号  Date issued：2017.9

J-GLOBAL

researchmap

Applicant：日本放送協会

Application no：特願2013-132070  Date applied：2013.6

Announcement no：特開2015-008367  Date announced：2015.1

J-GLOBAL

researchmap

Applicant：日本放送協会

Application no：特願2013-132070  Date applied：2013.6

Announcement no：特開2015-008367  Date announced：2015.1

Patent/Registration no：特許第6200220号  Date issued：2017.9

J-GLOBAL

researchmap

Applicant：日本放送協会

Application no：特願2012-282069  Date applied：2012.12

Announcement no：特開2014-127788  Date announced：2014.7

J-GLOBAL

researchmap

Applicant：日本放送協会

Application no：特願2012-282070  Date applied：2012.12

Announcement no：特開2014-127789  Date announced：2014.7

J-GLOBAL

researchmap

Applicant：日本放送協会

Application no：特願2012-283282  Date applied：2012.12

Announcement no：特開2014-126694  Date announced：2014.7

J-GLOBAL

researchmap

Applicant：日本放送協会

Application no：特願2012-282070  Date applied：2012.12

Announcement no：特開2014-127789  Date announced：2014.7

Patent/Registration no：特許第6076083号  Date issued：2017.1

J-GLOBAL

researchmap

Applicant：日本放送協会

Application no：特願2012-282069  Date applied：2012.12

Announcement no：特開2014-127788  Date announced：2014.7

Patent/Registration no：特許第6076082号  Date issued：2017.1

J-GLOBAL

researchmap

Applicant：日本放送協会

Application no：特願2012-283282  Date applied：2012.12

Announcement no：特開2014-126694  Date announced：2014.7

Patent/Registration no：特許第6080541号  Date issued：2017.1

J-GLOBAL

researchmap

Applicant：日本放送協会

Application no：特願2012-190032  Date applied：2012.8

Announcement no：特開2014-049875  Date announced：2014.3

J-GLOBAL

researchmap

Applicant：日本放送協会

Application no：特願2012-190032  Date applied：2012.8

Announcement no：特開2014-049875  Date announced：2014.3

Patent/Registration no：特許第5937926号  Date issued：2016.5

J-GLOBAL

researchmap

Applicant：日本放送協会

Application no：特願2012-131787  Date applied：2012.6

Announcement no：特開2013-258465  Date announced：2013.12

J-GLOBAL

researchmap

Applicant：日本放送協会

Application no：特願2012-131787  Date applied：2012.6

Announcement no：特開2013-258465  Date announced：2013.12

Patent/Registration no：特許第6195404号  Date issued：2017.8

J-GLOBAL

researchmap

Applicant：日本放送協会

Application no：特願2011-233110  Date applied：2011.10

Announcement no：特開2013-093668  Date announced：2013.5

J-GLOBAL

researchmap

Applicant：日本放送協会

Application no：特願2011-231113  Date applied：2011.10

Announcement no：特開2013-090253  Date announced：2013.5

J-GLOBAL

researchmap

Applicant：日本放送協会

Application no：特願2011-231113  Date applied：2011.10

Announcement no：特開2013-090253  Date announced：2013.5

Patent/Registration no：特許第5789172号  Date issued：2015.8

J-GLOBAL

researchmap

Applicant：日本放送協会

Application no：特願2011-226230  Date applied：2011.10

Announcement no：特開2013-090015  Date announced：2013.5

J-GLOBAL

researchmap

Applicant：日本放送協会

Application no：特願2011-226230  Date applied：2011.10

Announcement no：特開2013-090015  Date announced：2013.5

Patent/Registration no：特許第5898452号  Date issued：2016.3

J-GLOBAL

researchmap

Applicant：日本放送協会

Application no：特願2011-220377  Date applied：2011.10

Announcement no：特開2013-081100  Date announced：2013.5

J-GLOBAL

researchmap

Applicant：日本放送協会

Application no：特願2011-220377  Date applied：2011.10

Announcement no：特開2013-081100  Date announced：2013.5

Patent/Registration no：特許第5841395号  Date issued：2015.11

J-GLOBAL

researchmap

Applicant：日本放送協会