Video data stalling detection method and system

Abstract

The present invention relates to the technical field of detection, and particularly relates to a video data stalling detection method and system. The video data stalling detection method comprises the steps of acquiring to-be-detected video data, and splitting the video data frame by frame so as to obtain multiple frames of picture data; sequentially storing the multiple frames of picture data in a queue according to an actual time sequence of all frames of picture data, sequentially calculating similarity of every two frames of picture data, if the similarity is greater than a preset similarity threshold, marking the two frames of picture data similar, counting the frames of picture data that are similar, if two frames of picture data are not similar, determining whether the similarity quantity is within a preset similarity quantity threshold, so as to determine whether the data is stalled, and if the data is stalled, recording stalled picture data in a stalling list. The stalled picture data caused by frame loss can be found in the stalling list, and the specific lost frame and a video time point of the specific lost frame can be known.

Description

technical field [0001] The invention relates to the technical field of detection, in particular to a video data freeze detection method and system. Background technique [0002] At present, most video data freeze detection methods use real-time acquisition of the current FPS (frames per second) value of the device, and judge the difference. If the FPS value drops sharply at a certain moment in the same scene, and then returns to normal, then It is believed that there is a Caton phenomenon at this time. [0003] However, using the FPS value to judge Caton has the following disadvantages: [0004] 1. In some cases, the freeze phenomenon is caused by frame loss. In this case, the FPS difference can be used to judge whether it is freeze; in some cases, the freeze is caused by the extremely uneven and uneven Reasonably measured time intervals and incorrect processing of the buffered frame number, etc. The frame rate at this time is normal, so it is impossible to judge the stutt...

AI Technical Summary

Problems solved by technology

[0004] 1. In some cases, the freeze phenomenon is caused by frame loss. In this case, the FPS difference can be used to judge whether it is freeze; in some cases, the freeze is caused by the extremely uneven and uneven Reasonably measured time intervals and incorrect handling of buffered frames, etc. The frame rate at this time is normal, so it is impossible to judge the stuttering phenomenon through the FPS value

[0005] 2. The system platforms are different, and the methods or ways to obtain the FPS value are also different. Using the FPS difference to judge whether there is a freeze cannot be applied to the freeze detection of multiple different platforms at the same time.

[0006] 3. Use the FPS difference to judge whether it is stuck. It is only judged by data comparison. The screenshot or video of the scene at the time of the freeze is not retained. If you add a screenshot or video save function, there will be a certain amount of equipment performance due to such functions. Causing the scene to freeze, thereby reducing the credibility of the test results; if external equipment such as a camera is used for recording, it is susceptible to interference from environmental factors (such as light and dark), which may cause the situation that the freeze scene cannot be truly presented

Images