Method and apparatus for video quality diagnosis based on scene classification

Abstract

The invention is applicable to the technical field of video quality diagnosis, and provides a method and apparatus for video quality diagnosis based on scene classification. The method includes: obtaining real-time image data; classifying scenes of the real-time image data by employing a trained scene classifier to obtain the scene category and the category confidence degree corresponding to the real-time image data; obtaining a diagnosis item corresponding to the scene category and the diagnosis item sensitivity according to the corresponding relationships between the scene category, the diagnosis item and the diagnosis item sensitivity if the category confidence degree is greater than or equal to a preset threshold; and adjusting the diagnosis item of a video quality diagnosis system toreach the diagnosis item sensitivity in order to perform the video quality diagnosis of the real-time image data. According to the method and apparatus, the diagnosis item sensitivity of the video quality diagnosis system can perform automatic adjustment according to the scene category and the category confidence degree of the real-time image data, and the problems of false report and report failure of the video quality diagnosis can be solved.

Description

technical field [0001] The invention belongs to the technical field of video quality diagnosis, and in particular relates to a method and device for video quality diagnosis based on scene classification. Background technique [0002] In the video quality diagnosis, some diagnostic items are often diagnosed, for example, the diagnosis of video color cast, sharpness, abnormal brightness, black screen, jitter, freezing, noise and other diagnostic items. When these diagnostic items are diagnosed, it will be Related to alarm sensitivity. The alarm sensitivity refers to the critical index for judging abnormal situations such as color cast, sharpness, abnormal brightness, black screen, jitter, freezing, and noise in the video image. When the alarm sensitivity is higher, it is easier to trigger an alarm, and the probability of false alarms for abnormal conditions such as color cast, clarity, abnormal brightness, black screen, jitter, freezing, and noise in the video image increases...

AI Technical Summary

Problems solved by technology

[0004] In view of this, the embodiment of the present invention provides a method and device for video quality diagnosis based on scene classification, so as to solve the problem of false positives and false positives that cannot be solved in the prior art of video quality diagnosis

Images