HVS (Human Visual System)-based objective quality evaluation method for three-dimensional (3D) information

Abstract

The invention belongs to the field of image processing, provides a method for evaluating the quality of compression distortion three-dimensional (3D) images and particularly provides an HVS (Human Visual System)-based objective quality evaluation method for 3D information, and the HVS-based objective quality evaluation method has better result consistency with subjective evaluation. The technical scheme adopted by the invention is as follow: the HVS-based objective quality evaluation method comprises the following steps of: (1) converting an original 3D color image used in an experiment into a gray level image; (2) processing the converted gray level image by adopting a discrete Contourlet transforming filter group; (3) obtaining the decomposition coefficients of 3D image pairs in different spatial frequency bands after decomposition by the discrete Contourlet transforming filters; (4) realizing a masking effect through carrying out gain control on data which are subjected to CSF (Critical Success Factor) weighting in the spatial frequency bands; (5) obtaining 3D viewpoint pairs which are subjected to HVS modeling; and (6) evaluating the objective quality of the 3D image by adopting HPSNR (High Peak Signal-Noise Ratio). The HVS-based objective quality evaluation method for the 3D information is mainly applied to the design and manufacture of image sensors.

Description

technical field [0001] The invention belongs to the field of image processing, and is aimed at an objective quality evaluation system of stereo images, in particular to an HVS-based objective quality evaluation method of stereo information. Background technique [0002] At present, the quality evaluation methods of digital plane images can be divided into two categories: subjective evaluation and objective evaluation. The subjective quality evaluation process needs to select a large number of observers, and each observer has to conduct a large number of repeated experiments on multiple test images, which is time-consuming and expensive, and is susceptible to the complexity and differences of the individual psychological and physiological changes of the testees. Influence, the operation is difficult and cannot achieve real-time performance. Therefore, how to establish an objective quality evaluation method consistent with the subjective evaluation results has become an urgen...

AI Technical Summary

Problems solved by technology

However, since PSNR, like MSE, only considers the difference between pixels and does not consider the visual characteristics of the human eye, this evaluation method also estimates the difference between the original image and the distorted image as a whole, and cannot reflect the difference in the local area of ​​the image. different situations

Images