A method for checking the correctness of multimedia optimization

Abstract

The invention discloses a method for detecting the correctness of multimedia optimization. The method includes the following steps: inputting a media stream sample into a public processing module, after which a tee element is inserted into the public processing module, and the two outputs of the tee element are the same One of the video streams flows into the pre-optimized processing module, the other into the optimized processing module, and the two data signals flow into the comparison module for comparison and verification. If the verification result is correct, it flows into the lower-level processing element, and if the verification result is incorrect, an error is reported. The method of the invention is simple, and when the two comparison results are shown to be consistent, the correctness of the manually compiled assembly code can be verified and can be used continuously. Moreover, the verification method does not require manual participation, which saves a lot of labor costs, and the transmission process is simplified and stable, avoiding errors that may be caused by a large amount of redundant data transmission.

Description

technical field [0001] The invention relates to a signal transmission optimization verification method, in particular to a multimedia optimization correctness detection method. Background technique [0002] Now, traditional multimedia signal transmission process optimization is usually achieved by optimizing hot functions, that is, using manually written assembly codes to replace the codes automatically generated by the original compiler, so as to obtain a leap in performance of hot functions. The reason is that the code automatically generated by the compiler cannot run as efficiently as the code carefully written by hand. [0003] However, although the assembly code written manually is much more efficient than the code automatically generated by the compiler, the writing process is extremely cumbersome and error-prone. How to verify that the optimized function is functionally consistent with the code automatically generated by the compiler before optimization has always b...

AI Technical Summary

Problems solved by technology

The reason is that the code automatically generated by the compiler cannot run as efficiently as the code written carefully by hand

[0003] However, although the assembly code written manually is much more efficient than the code automatically generated by the compiler, the writing process is extremely cumbersome and error-prone

How to verify that the optimized function is functionally consistent with the code automatically generated by the compiler before optimization has always been a relatively difficult step

At present, more human eye observation methods are used, that is, to observe whether the output image or video after the optimization of the multimedia link is distorted, and to judge whether there is a page missing. This not only requires manual participation, but also increases labor costs, and is very imprecise.

Images