MOTION COMPENSATING PREDICTIVE ENCODING PART

摘要

PURPOSE:To decrease the redundancy between fields and make the encoding efficiency excellent by encoding a frame error signal according to the correlation between picture elements in the signal. CONSTITUTION:A motion detection part 11 compares the image signal of a block to be encoded in an input image, consisting of data in picture element units, with an encoded image signal and detects a motion vector from the similarity between the signals. According to the detected motion vector, a subtracter 3 subtracts a field composite motion prediction signal 201 read out of a memory 1 from a field composite frame input signal 200 to find a field composite prediction error signal 110. A field composition part 10 puts error signals of respective fields together and supplies the result as the frame error signal 110 to an encoding part 4. The encoding part 4 encodes the signal 110 according to the correlation between the adjacent picture elements in the signal. Then inter-frame prediction is performed to remove spatial redundancy from the composite frame, so successive correlation can be utilized and the encoding efficiency is improved.