A STUDY ON VIDEO SCRAMBLING CONSIDERING INTER-FRAME PREDICTION(International Workshop on Advanced Image Technology 2006)
スポンサーリンク
概要
- 論文の詳細を見る
This paper considers the scrambling method for MPEG-2 and other similar video codecs that exploit the inter-frame prediction feature. Usually this kind of methods scrambles the data (e.g., DCT coefficients) in I-frames, and other frames (P-/B-frames) can be automatically scrambled due to the inter-frame prediction. This is efficient in reducing the computational complexity. However, it is known that intra-encoded macroblock (intra-MB) may exist in P-frames, which de-scrambles the MB and weakens the scrambling strength. In this paper we investigate the relationship between the number of intra-MBs and the scrambling effects in P-frames, and based on it, we propose a scrambling scheme which scrambles several frames instead of all frames. It is justified through analysis of distribution of local PSNR and its relation to subjective picture quality.
- 社団法人電子情報通信学会の論文
- 2006-01-02
著者
-
SAKAZAWA Shigeyuki
KDDI R&D Labs. Inc.
-
Zhao Liang
Faculty Of Engineering Utsunomiya University
-
NEMOTO Yoshio
Faculty of Engineering, Utsunomiya University
-
TOYOTA Yousuke
KDDI R&D Laboratories Inc.
-
YAMAMOTO Hideo
Faculty of Engineering, Utsunomiya University
-
Nemoto Yoshio
Faculty Of Engineering Utsunomiya University
-
Toyota Yousuke
Kddi R&d Laboratories Inc.
-
Sakazawa Shigeyuki
Kddi R&d Laboratories Inc.
-
Yamamoto Hideo
Faculty Of Engineering Utsunomiya University
-
Zhao Liang
KDDI R&D Laboratories Inc.
関連論文
- Light Weight MP3 Watermarking Method for Mobile Terminals
- Video Data Transmission Protocol "SVFTP" Using Multiple TCP Connections and Its Application(Networks)
- Proposal of A Supporting System for Taking Care of Physically Disabled Peopl using Sound and Image Processing ((放送方式、放送現業、無線・光伝送)2000 Asia-Pacific Symposium on Broadcasting and Communications)
- A STUDY ON VIDEO SCRAMBLING CONSIDERING INTER-FRAME PREDICTION(International Workshop on Advanced Image Technology 2006)
- D-20-5 MODELING HUMAN MOTIONS SUCCESSIVELY IN MULTIPLE LATENT SPACES
- Efficient Generation of Dancing Animation Synchronizing with Music Based on Meta Motion Graphs
- Optimized Codebook Construction and Assignment for Product Quantization-based Approximate Nearest Neighbor Search
- Large-scale Image Retrieval as a Classification Problem
- Optimized Codebook Construction and Assignment for Product Quantization-based Approximate Nearest Neighbor Search
- Large-scale Image Retrieval as a Classification Problem
- 3-3 Multimodal Synchronization for Virtual Agents
- 3-6 Cooperative Segmentation of Human Motions for Affective Agents