| 摘要 |
In a video coding device capable of making coded data have a hierarchical structure, a specified area of each frame is selected, the position and the shape of the selected area are encoded, a pixel value of the selected area is encoded as lower-layer coded data, a pixel value of a whole image is encoded as first upper-layer coded data by using pixel values of already decoded images of the lower-layer and the first upper layer and a pixel value of the selected area is encoded as second upper-layer coded data by using pixel values of already decoded images of the lower-layer and the second upper layer. By the video coding device and video decoding device mentioned above, decoding a part of coded data makes it possible to reproduce only a selected area of a lower image-quality or reproduce a whole image of a lower image-quality or reproduce a whole image with a selected area of a higher image-quality and other areas of a lower image- quality. In encoding and decoding a background video-sequence and a plurality of foreground part-video- sequences, a coding device (101, 102, 103, 122, 123, 120) encodes position information of part-video-sequences and a decoding device (105, 106, 107, 108, 109, 110, 124, 125, 121) prepares weight values necessary for weighted-mean image synthesizing on the basis of the coded data of the position information data, whereby each part-video-sequences is synthesized with the background video-sequence by using the weighted mean values. And in laying a plurality of part-video-sequences over a background video-sequence by using weighted mean values, an amount of coded data can be reduced because the weighted mean values are prepared from binarized information. Furthermore, a synthesized video-sequence has a smoothed boundary between the part images and the background image can be smoothly synthesized without any visual defect since the weighted mean data prepared from the binarized information gets a value of 0 to 1. In synthesizing an lower-layer frame from preceding and proceeding lower-layer frames by using a first part-area-information and a second part-area-information of lower-layer frames existing temporally before and after the synthesizable frame, the temporally preceding and proceeding lower-layer frames previously averaged with weight is used for synthesizing an overlapped portion of a first part-area with a second part-area or an area not belonging to the first part-area and the second part-area on the synthesizable frame, the temporally proceeding lower-layer frame used for synthesizing an area belonging to the first part-area only and the temporally preceding lower-layer frame is used for synthesizing an area belonging to the second part-area only. By the video coding device and video decoding device with the synthesizing method as above described, the synthesyzing image is free from the distortion occurred in the prior art image aforementioned. <IMAGE> |
