| 摘要 |
PROBLEM TO BE SOLVED: To eliminate the need for use of a plurality of search blocks by converting image data from one of the time domain or the frequency domain into the other, generating a motion vector in one domain and converting the vector into the other domain. SOLUTION: A succeeding image in a picture group GOP to be compressed is in general a P(predictive) picture based on an I(intra-) picture and fed to a motion estimate device 30 from a frame recorder 10, in which a motion vector denoting a motion of image between the I and P pictures is generated. A motion prediction device 40 uses the motion vector and the decoded I picture to generate a predicted P picture. The predicted P picture is subtracted from the actual P picture at a subtractor 50 and the difference between the two frames is fed to a wavelet transformation unit 80, in which the data are compressed. The compressed difference data are outputted from an entropy coder 100 and decoded by a decoding circuit and the difference data are generated again. Thus, the data are compressed based on an image obtained by the decoding circuit. |
