IMAGE SIGNAL CODING METHOD AND DEVICE, IMAGE COMPLEXITY PREDICTION METHOD AND DEVICE THEREOF

摘要

PROBLEM TO BE SOLVED: To encode an image without deteriorating the quality of the coded image by using a reference generated bit amount to decide a quantization scale used for coding the image and using the decided quantization scale to encode the image for each macro block. SOLUTION: A reference generating bit amount calculation circuit 5 calculates both a reference generating bit that is predicted from a motion vector signal MV and a reference generating it that is predicted from a mean value separation residue with respect to each macro block MB of a received input MB signal S10. Since the image signal coder 1 discriminates whether the difficulty of coding is predicted, based on a reference generating bit amount in the inter coding mode or based on a reference generating bit amount in the intra-coding mode for each macro block MB in an image, prediction of coding difficulty is switching for each macro block. Thus, the quantization scale is stably controlled for conducting coding.