An evaluation method and evaluation system for stereoscopic video quality

Abstract

The invention discloses a stereo video quality evaluation method and a stereo video quality evaluation system. The method is simple in operation, low in cost and good in transplantable performance. According to the method and the system, the accuracy of the evaluation results is high; the stereo video quality can be effectively judged; the stereo videos with high qualities can be screened out; the high quality videos are provided for users by industries; the qualities of various stereo video compression, transmission, denoising and watermark embedding methods can be effectively judged; improvement and optimization of stereo video compression algorithm, transmission mode and watermark algorithm can be instructed; the features of the human eye visual system are fitted; and the evaluation results are more approximate to the watching feelings of people.

Description

technical field [0001] The invention relates to the field of quality evaluation of stereoscopic video, in particular to the evaluation of depth information of stereoscopic video. Background technique [0002] Stereoscopic video is becoming more and more popular, and people watching stereoscopic video can obtain the feeling of being on the scene personally, and this is because stereoscopic video can not only provide content information, but also can provide the depth information of scene. The depth information of stereoscopic video is reflected by the absolute distance between people and objects and the relative distance between objects. Among them, the judgment of the relative distance between objects is more important, which will directly affect the perception of depth information. Perceived depth information enables people to obtain a better visual experience, but operations such as compression, transmission, and watermarking of stereoscopic video will reduce video quality...

AI Technical Summary

Problems solved by technology

Since factors such as the evaluator's preference, state, emotion, test time, and knowledge background will affect the evaluation results, the results obtained through subjective evaluation methods contain many uncertain factors, which cannot accurately explain the quality of stereoscopic video.

Moreover, this method consumes huge manpower and material resources, has poor portability, and cannot meet the needs of real-time applications, so it cannot be applied in engineering

[0005] Objective evaluation creates evaluation indicators including brightness contrast distortion, structural similarity and depth fidelity by extracting stereoscopic video features, and then uses regression analysis to determine the weight of each evaluation indicator’s contribution to stereoscopic video quality, and then obtains the evaluation indicators and The mathematical model between video quality, so the process of this method is very complicated, and some information will be discarded in the process of making and transmitting stereoscopic video, resulting in video distortion, so the extracted video features are not comprehensive and accurate, so the objective evaluation results cannot be effective To accurately judge the quality of stereoscopic video

Images