Robust mechanism research method of feature saliency in image quality assessment

Abstract

The invention discloses a robust mechanism research method of feature saliency in image quality evaluation. The method steps are as follows: firstly, establish the objective function of feature selection in image quality evaluation and initialize model parameters; secondly, select the optimal feature and add The feature matrix and remove feature interference items; then, calculate the significance of feature selection in the image quality evaluation system; then, judge whether the robustness requirements of the system are met or the upper limit of the number of features is reached; finally, verify the classification effect of the model. The present invention measures feature salience by introducing system feature signal-to-noise ratio, solves the constrained optimization problem of smooth convex function in the image quality evaluation system, effectively reduces the interference of insignificant features on the classification surface, and improves the robustness of the image quality evaluation system. Stickiness, which solves the adaptive optimization problem of feature attribute selection in the image quality evaluation network based on learning mechanism.

Description

technical field [0001] The invention relates to the field of computer vision research, in particular to a robust mechanism research method of feature saliency in image quality evaluation. Background technique [0002] Image quality is an inherent attribute of an image, and is generally obtained by measuring the degree of image degradation. Image quality evaluation is a way to measure the degree of image degradation, which has wide application value in the fields of image processing, computer vision and system engineering. The application of the method has theoretical and practical significance. [0003] So far, there has not been a unified quality evaluation standard in the field of image restoration, and evaluation methods are usually divided into subjective evaluation methods and objective evaluation methods. Among them, the subjective evaluation method is easily affected by the observer's knowledge background, psychological motivation and other factors, and cannot be em...

AI Technical Summary

Problems solved by technology

The limitations of this type of method are often reflected in: in the case of fewer samples, the system cannot adaptively select reasonable feature attributes according to different evaluation objectives

That is to say, the evaluation system cannot screen the appropriate feature input according to the actual situation, and the selection of insignificant features into the system often leads to overfitting in the learning process.

Images