Image coder and coding method

Abstract

When making a block match, the method comprises: in the first, reading the reference window in a reference frame; the location of the reference window is called the first location; then, reading the macro-block in two frames to be estimated, the location of the macro-block is called the second location; wherein, the second location is corresponding to the first location; then, executing the match between the macro- block and the reference window; finally, changing the reading location of the first and second locations, and repeating above steps until completing the macro-block match of all macro-blocks of at least two frames in the frame to be estimated.

Description

technical field [0001] The present invention relates to an image coding method, and in particular to a video image coding device and device suitable for having multiple reference frames (reference frames, referred to as reference frames, hereinafter referred to as reference frames) and multiple frames to be estimated. method. Background technique [0002] The data volume of digital video images is usually extremely large. When performing digital image encoding or image compression, in addition to requiring huge storage capacity and high-speed digital operation speed, a large amount of data transmission bandwidth is also required. Therefore, traditionally, when encoding or compressing higher-quality frames, higher memory bandwidth is required, which indirectly affects the transmission speed of other data in the system. [0003] For the animation compression standard (Motion Picture Experts Group, MPEG), video image encoding or image compression usually utilizes three types o...

AI Technical Summary

Problems solved by technology

In the case of limited memory bandwidth, repeated reading of a large amount of image data will cause a burden on the system memory bandwidth, which in turn will affect the efficiency of other work in the system

Images