Method for mapping image address in memory

Abstract

The method is used for mapping the decoded image into the relevant address of the memory. It comprises: dividing one image frame into multi rectangle macro blocks; providing a memory module for saving the image data; said memory module at least comprises two memory arrays; each storage array comprises multi memory rows; using the integer number of adjacent micro blocks in one image frame as one mapping unit; using one mapping unit as an unit, according to the sequence from the left to the right in horizontal direction, and then according to the sequence from the top to the bottom in vertical direction, sequentially mapping the image data to the memory, and until the memory rows is full; switching over the memory arrays in order to map the image data in next mapping unit, until completing the mapping of one image frame.

Description

technical field [0001] The invention relates to an image storage technology, including a storage technology of video image data in a synchronous dynamic random access memory (SDRAM), in particular to an address mapping method of an image in the memory. Background technique [0002] Digital video transmission technology is widely used in multimedia fields, such as digital television (HDTV), set-top box, DVD, etc. In order to reduce the bandwidth requirement, methods such as encoding and decoding are often used for video compression. Digital video compression algorithms typically reduce the content information of an image or group of images by eliminating redundant components present in the video signal. For example, in MPEG2, MPEG4, H.264 and other video compression algorithms, a frame of image is divided into multiple 16*16 or 8*8 macroblocks to eliminate redundant components in space; at the same time, motion prediction and motion compensation are adopted Eliminate redund...

AI Technical Summary

Problems solved by technology

There is a big difference in the delay time between the two

[0006] However, the starting address of the reference macroblock input by the reference frame in motion compensation is uncertain, which may lead to frequent access to the memory, which may cause frequent pre-registration of the memory Charge

The bandwidth of the reference macroblock in the MPEG algorithm is generally more than 2 / 3, especially when the bit width of the memory increases, the bandwidth of the reference macroblock may further increase

Images