A novel lossless coding scheme for video signals is described. The scheme uses a block-adaptive 3D prediction method that predicts a video signal based on both the current and motion-compensated previous frames. The resulting prediction errors are encoded by using a kind of context-adaptive arithmetic coder. Several coding parameters such as prediction coefficients and motion vectors are iteratively optimized for each frame so that the overall coding rate can be a minimum. Moreover, a quadtree-based variable block-size motion compensation technique is introduced for efficient representation of motion information. Experimental results show that the coding rates of our scheme are 11-17% lower than those of the H. 264-based lossless coding scheme.
We propose a robust method of shot change detection for old film sequences. In particular, we propose a cut detection method using phase-only correlation. The threshold value of this method can be determined mathematically using the properties of the phase-only correlation in the proposed method. Dissolves can be detected by improving the proposed cut detection method. We evaluated the proposed methods by comparing it to several conventional shot change detection methods on actual old film sequences. Experimental results show that the proposed method can detect shot changes in old film sequences more accurately than conventional methods.
The counterfeiting and illegal use of ID cards have become problems as the cards are used more in society. To prevent the problems, systems that authenticate ID cards by embedding digital watermarks in the facial images of the cards have been researched. In previous research, the facial images of the cards were taken with scanners, but our system uses a video camera. A user shows an ID card in front of the video camera, and the facial image on the card is extracted from the motion picture. After the card shape is corrected, a watermark is detected. To detect a watermark from a motion picture shot by the video camera, the robustness and the calculation processing time of a previous watermarking method using DFT and spread spectrums were improved. An experiment to detect watermarks from ID cards was conducted, and the utility of this system was shown.
The presence (i.e. sensation of being there) from a video system is well known to increase as the viewing angle increases. Unfortunately, quantitative evaluation of presence is very difficult. Previous evaluations attempted to determine both subjective and objective assessments. We quantified presence in viewing still images by varying the video system viewing angle along with subjective evaluations using well selected matched words and objective measurements of viewers' body sway. We also examined the correlation between the indices used to quantify presence. The results indicated that the correlation between the evaluation words “presence” and “powerfulness” was comparatively high but that it was almost zero between “presence” and the total distance of viewers' body sway. This suggested that subjective evaluations using the words “presence” and “powerfulness” were both valid to assess presence and that the subjective evaluations of presence did not correlate with the body sway for each experiment trial.
We developed and evaluated the e-Life Amenity System, a system that provides amenities to elderly people and children. From the view point of operability, the evaluation indicates that children preferred high degrees of input freedom. Furthermore, elderly people preferred simple 2D video to 2D video in 3D cyberspace, the exact opposite of children. From the view point of avatar design, the results showed that both elderly people and children preferred avatars that had facial images when the communication partner was a member of the family. When the partner was not a member of the family, elderly people still preferred the use of facial images, while children preferred avatars without facial images. Furthermore, both elderly people and children preferred simple 2D facial images to avatars having 2D block heads with facial images when the communication partner was a member of the family.
Perceptual segmentations with illusory stratifications were studied by using silhouettes each composed of a circle and a figure having four sharp vertices. The silhouettes were perceptually segmented into the circular areas and the areas with four vertices. Circles are well known as the most typical good figures. We studied whether such a good figure was dominantly perceived as being in front of competitive segmentations. The length of illusory contours is a factor that makes perceptual segmentation difficult. Areas other than common ones in competitive segmentations are factors contributing to perceptual segmentations. The boundary conditions for competitive segmentations in stratification were clarified in terms of both the relative lengths of illusory contours and the relative sizes of areas other than the common ones. Experimental results showed that the areas with the sharp vertices were dominantly perceived as being in front of the circular areas.