No-reference stereo image quality evaluation method based on a quality map generation network

Abstract

The invention discloses a no-reference stereo image quality evaluation method based on a quality map generation network. The method comprises the steps of in a training stage, constructing a quality map generation network, sequentially inputting normalized images of all distorted plane images in a training set into the quality map generation network, taking a corresponding quality map as supervision, and performing training to obtain an optimal quality map generation network model; Test phase, utilizing the optimal quality map to generate a network model to extract a left viewpoint image of the distorted stereo image to be evaluated; predicting quality maps of the normalized images of the right viewpoint image and the fused image respectively, acquiring respective salient feature maps, solving respective predicted quality scores according to the corresponding predicted quality map and the corresponding salient feature map, and fusing the three predicted quality scores to obtain a finalpredicted quality score; The method has the advantages that the influence of various characteristics of the stereoscopic image on the visual quality can be fully considered, so that the correlation between an objective evaluation result and binocular subjective perception can be improved.

Description

technical field [0001] The invention relates to an image quality evaluation technology, in particular to a no-reference stereo image quality evaluation method based on a quality map generation network. Background technique [0002] With the rapid development of stereoscopic image applications, many related stereoscopic image technologies and services have been introduced into people's daily life and many professional fields. Various distortions may occur during the collection, transmission, processing and display of stereoscopic images. Therefore, it is of great practical significance to establish a high-performance stereoscopic image quality evaluation method. Stereoscopic image quality evaluation can be divided into two ways: subjective evaluation and objective evaluation. The subjective evaluation is that people directly evaluate the stereoscopic image subjectively. Since the Human Visual System (HSV) is the final receiver of the stereoscopic image, the subjective evalu...

AI Technical Summary

Problems solved by technology

[0003] Most of the existing stereoscopic image quality evaluation methods cannot accurately judge the quality of stereoscopic images, and rarely consider the versatility of the model. They are all trained and tested on a database, and image saliency is not considered as an object.

The weight distribution of the left and right viewpoints of stereoscopic images also seriously affects the quality of stereoscopic images, and is rarely considered in the existing stereoscopic image quality evaluation methods.

Images