| 摘要 |
<p>PROBLEM TO BE SOLVED: To reduce code amount generated by converting composited images to the same resolution as those of the respective frame images of an area under consideration and a background, turning them into reference image data and predicting the respective movement compensation of both frame images. SOLUTION: Input image data 101 are separated into an image data area under consideration 103 and background area image data 1041 in an area separation part 102. The data 1041 are converted into low resolution background area image data 1042 in a resolution conversion part 105. The image data area under consideration 103 and the image data 1042 are resolution converted into a multiplexing part 106, then combined into one frame image, multiplexed to encoding object image data 107 and converted. After they are encoded in an encoding part 108, encoded data 109 are outputted. Also the multiplexing part 106 sends data, indicating whether a macroblock to be encoded belongs to the area under consideration or belongs to a background area, to an area kind deciding part 125 as area arrangement data 126. At that time, a part equivalent to the area under consideration in the background area is painted full as white, bit rate is reduced, and processing loads are reduced.</p> |
