Stereo image objective quality evaluation method based on integration of visual threshold value and passage

Abstract

The invention discloses a stereo image objective quality evaluation method based on integration of a visual threshold value and a passage. The stereo image objective quality evaluation method based on the integration of the visual threshold value and the passage includes: conducting visual threshold value analysis and singular value decomposition by distributing passages on an a left viewpoint image of an undistorted stereo image and a right viewpoint of the undistorted stereo image and a left viewpoint image of a distorted stereo image and a right viewpoint of the distorted stereo image; conducting visual threshold value analysis and singular value decomposition by distributing passages on difference images of the left viewpoint image of the undistorted stereo image and the right viewpoint of the undistorted stereo image and the left viewpoint image of the distorted stereo image and the right viewpoint of the distorted stereo image, and obtaining a stereo perception evaluating measurement by utilizing a liner combination method; finally conducting linear weighting combination on an objective evaluating scale measurement and a stereo perception evaluating measurement to obtain an overall quality evaluating objective value of the stereo image. The stereo image objective quality evaluation method based on the integration of the visual threshold value and the passage not only effectively utilizes visual properties such as a just-noticeable distorted visual threshold value, a visual covering effect and stereo perception, but also evaluates the stereo perception on the basis of evaluating the left viewpoint image quality and the right viewpoint image quality and improves relevance of an objective quality evaluating result and subjective perception.

Description

technical field [0001] The invention relates to a method for evaluating the quality of a stereoscopic image, in particular to a method for evaluating the objective quality of a stereoscopic image based on visual threshold and channel fusion. Background technique [0002] Stereo image, as a new type of image that can provide a sense of stereo, has become an important direction of 3D research, and has been widely concerned and applied. Compared with traditional multimedia, stereoscopic video can create a more intuitive and realistic scene experience for the public, but it needs to process at least twice as much data. Therefore, while improving the encoding and compression efficiency of stereoscopic images, it is also necessary to ensure the subjective perception of stereoscopic images. Because the subjective evaluation method of stereoscopic image quality is not only easily affected by various factors such as objective conditions, subjective emotions, and the observer's own c...

AI Technical Summary

Problems solved by technology

At present, many researchers use planar image quality evaluation methods directly for the evaluation of stereoscopic images, such as quality evaluation based on singular value decomposition and evaluation based on structural similarity. The effect is not very ideal, because they do not consider the human visual characteristics and the depth perception characteristics unique to stereoscopic images, resulting in the final objective evaluation and the subjective evaluation of the human eye are not very consistent.

Images