Video playback test method, device, electronic equipment and storage medium

Abstract

The invention discloses a video playing test method and device, electronic equipment and a memory medium. The method comprises the steps of determining a target video player; calling a plurality of operation API interfaces of the target video player to establish a plurality of scenes, wherein each operation API interface is used for indicating that each operation behavior is packaged into an operation interface in advance according to a target standard; controlling the target video player to play a to-be-tested video according to a plurality of simulation scenes; carrying out screenshot on a currently played image of the to-be-tested video according to a preset frequency in a process of playing the to-be-tested video; and carrying out playing test on the to-be-tested video according to a pre-established abnormal image quality detection model and the obtained screenshot image. According to the method, the decoupling of players can be supported, various complex interaction operation scenes in the playing process can be realized, whether the video playing is abnormal or not can be judged through screenshot and deep learning of the abnormal image quality, the artificial examination cost is greatly reduced, and the efficiency is improved.

Description

technical field [0001] The present invention relates to the technical field of video processing, in particular to a video playback test method, device, electronic equipment and computer-readable storage medium. Background technique [0002] In recent years, video media has become more and more popular, and more and more Internet companies and super application providers have taken it as a strategic location. How to efficiently find quality problems in video playback has become an urgent problem to be solved and efficiency needs to be improved. [0003] In related technologies, the playback test of the video to be tested is usually realized mainly through automated tools such as Monkey Runner, that is, the automatic playback of the video is realized by encapsulating start, stop and automated test buttons. In terms of abnormality identification, it is possible to find out whether there is an abnormality by comparing the playback video frame with the local sample, and then real...

AI Technical Summary

Problems solved by technology

[0004] But, the problem that exists at present is, the above-mentioned mode by automation tool such as Monkey Runner, just for simple video operation (as start, stop etc.) The aspect does not support the decoupling of the video player, because different players may have different effects when playing the same video source, and this factor can easily affect the test results

In addition, for abnormal recognition, it is necessary to collect all video frames of the video source in advance. Currently, the minimum number of frames recognized by the human eye is 24 frames per second. Assuming a normal 120-minute video as an example, there are 17 pictures that need to be saved. More than 10,000 sheets, obviously this kind of consumption is very inefficient in the actual execution process

Images