| 摘要 |
PROBLEM TO BE SOLVED: To compress video data without deteriorating the picture of a video by generating difference data of the picture of a compressed picture group through entropy encoder, decoder, etc., and using it for compressing a next picture. SOLUTION: Compressed picture I data is sent from the entropy encoder 100 to an expansion chain consisting of the entropy decoder 110, an inverse quantization device 120 and an inverse wavelet transformation unit 130. Then one picture existing within the decoder is reconstituted and this is sent to a motion predicting device 40. Though the next picture of a picture group to be compressed GOP(group of picture) is a picture P by basing on the picture I as a reference, the picture is sent to a motion estimating device 30 from a frame reference number 10 to generate a motion vector showing the motion of a video within pictures I and P. The motion predicting device 40 generates predicted one of the picture P through the use of decoded one of motion vector picture I. Then, it is substrated from an actual picture P by a subtracter 50 to compress and what is compressed is decoded by expansion chains 110, 120 and 130 to generate difference data again. |
