image area most similar to the image region of the input image signal is discovered in the reference image from the frame memory 10 and motion vector is generated. From a reference image from the frame memory and the motion vector, the motion-compensated reference image is generated. By subtraction of the motion compensated reference image and the input image signal, the prediction residual is generated. Quantization and orthogonal transformation processing of the prediction residual and the processing by the result of the addition, the motion-compensated reference image and the inverse quantization and inverse orthogonal transformation process, a reference image is generated to be stored in memory. Orthogonal transform processing and quantization processing of the prediction residual to the variable length coding process by a video encoding apparatus which compression-coded video output signal is generated. The reference image is the inner side of the screen of the video display screen, the reference image A, B, C and the image display including the outer side of the screen outside the reference image D of the screen, and the screen outer reference image D in the reference image is the screen A, B, C of the a plurality of similar reference image is a, generated on the basis of the positional relationship of B. ;
展开▼
