Composite degraded image quality assessment method and system based on cartoon texture decomposition

Abstract

The invention belongs to the technical field of image assessment and discloses a composite degraded image quality assessment method and system based on cartoon texture decomposition. Firstly, significant edge structure information and texture detail information of an image are effectively distinguished with a cartoon texture decomposition algorithm; secondly, a cartoon component of the image is subjected to fuzzy assessment with a feature similarity algorithm, and fuzzy assessment factors are obtained; a noise intensity factor is obtained by use of the texture component; the factors are subjected to consistency correction, a weight parameter is given in combination with human eye vision characters, and a quality assessment model of a fuzzy noise composite degraded image is obtained finally. As for fuzzy and noise composite degraded image in an LIVEMD image library, the assessment result of the algorithm has higher subjective and objective consistency.

Description

technical field [0001] The invention belongs to the technical field of image evaluation, in particular to a method and system for evaluating the quality of composite degraded images based on cartoon texture decomposition. Background technique [0002] At present, the existing technologies commonly used in the industry are as follows: [0003] Images are the main source for people to obtain information. With the popularization of smart phones and digital cameras, a large number of images are acquired every moment. The rapid development of Internet technology makes the transmission and sharing of these images fast and convenient. In the process of image acquisition, transmission, processing and display, some unavoidable interference factors often lead to the degradation of image quality, such as electronic noise, jitter blur, data loss caused by compression, etc. Therefore, reliable quality evaluation of images It is the premise of accurate understanding and better use of im...

AI Technical Summary

Problems solved by technology

[0008] First, the quality degradation caused by blur and noise is the most common and unavoidable in the process of image acquisition, transmission and processing, and the image is easily affected by both at the same time, so exploring the composite degradation of the two has practical application value; and the prior art does not give good enlightenment;

[0009] Secondly, compared to blur and the compound degradation of JPEG and JPEG2000, the influence of blur and noise on the image features that the algorithm relies on to model has a reverse effect to some extent, and their mutual superposition and interference mechanisms are also more complicated.

[0010] In the prior art, there is no general fuzzy noise compound degraded image evaluation model. After the image is decomposed into a cartoon texture, the degree of blur and noise degradation is evaluated separately, effectively avoiding the mutual interference between the two;

[0011] Moreover, the existing technology does not obtain the final image quality measurement through the pooling strategy; resulting in weak subjective and objective consistency, it is impossible to accurately evaluate the composite degraded image quality

Images