Video quality assessment method based on visual salient regions and spatio-temporal characteristics

Abstract

The invention discloses a video quality evaluation method based on a visual saliency area and time-space characteristics. The shortcomings that the human visual characteristics and extracted features are not considered in the prior art, so that the video information cannot be completely reflected. The video quality evaluation method comprises the following steps: (1) extracting a video; (2) randomly selecting a frame of image from the selected video; (3) extracting the visual saliency area of the image; (4) judging whether all frames of images are extracted; (5) synthesizing a video; (6) obtaining a three-dimensional discrete cosine transform 3D-DCT coefficient; (7) extracting features; (8) reducing the feature dimension; (9) judging whether all videos are extracted; (10) predicting a mass fraction; (11) calculating related video coefficients; and (12) outputting the related coefficients. The video quality evaluation method has the advantages that the distortion type is not limited, the eye focus and time-space characteristics of the videos are fully considered, and that the evaluation result better satisfies the subjective evaluation result.

Description

technical field [0001] The invention belongs to the technical field of image and video processing, and further relates to a video quality evaluation method based on visually significant regions and spatio-temporal characteristics in the technical field of image and video quality evaluation. The invention can be applied to video used in video coding and video conferencing. According to the different influences of human eyes on images, salient areas of the video can be extracted, and the temporal and spatial characteristics of the video can be considered to evaluate the video objectively. Background technique [0002] With the rapid development of multimedia technology and computer networks, video signals are widely used in video surveillance, video conferencing and other services. When people come into contact with more and more videos, the demand for video-related services is also increasing. On the other hand, from video generation to transmission to end users, every stage ...

AI Technical Summary

Problems solved by technology

Although this method has added the HVS feature, it still has the disadvantage that the floating filter proposed by this method can only be implemented in the spatial domain, the calculation is quite complicated, and the spatio-temporal statistical characteristics of the video are not considered, so the evaluation results cannot be better in line with The results of subjective evaluation, and can only be used for compressed and distorted videos, and cannot be widely used in practice

However, the shortcomings of this method are that the extracted simple features cannot represent all the information of the video, and the impact of human visual characteristics on quality evaluation is not considered, and the evaluation results cannot better conform to the results of subjective evaluation.

Images