| 摘要 |
PURPOSE:To realize a three-dimensional block encoding without increasing the scale of a hardware, and to improve a compressing efficiency by discriminating removable blocks in an objective frame from the preceding and succeeding frames in a time-base direction against a frame to be processed, and removing the blocks. CONSTITUTION:The digital picture data of frame units inputted from an input terminal 1 are inputted to an encoding circuit 2, and the coefficient data of block units are inputted through frame memories 3 and 4 to a discriminating circuit 5. The circuit 5 discriminates whether or not the coma drop is valid based on each block DCT coefficient of a K frame and the before and behind K-1 K+1 frames in the time-base direction, and outputs a flag indicating that the coma is valid. The flag is transmitted to a block removal circuit 6, a calculating circuit 12, a switch 7, and an output terminal 1. The circuit 6 operates the coma drop, the DCT coefficients of the K, K-1, and K+1 frames are transmitted through a weighting circuit 8, and a delay circuit 9. quantized by a quantizing circuit 10, variable length-encoded by an encoding circuit 11, and outputted from a terminal 13. |
