Algorithm for evaluating comprehensive efficiency of objective method of image quality

Abstract

The invention discloses an algorithm for evaluating the comprehensive efficiency of an objective method of image quality. The algorithm has the following core: firstly, designing the performance evaluation index of a necessary objective evaluation algorithm of the image quality from different measurement angles so as to comprehensively evaluate the performance advantages and disadvantages of various algorithms; secondly, unifying the evaluation criterion of each index to cause the evaluation result data of each index to exhibit consistency; and finally, designing a simple and effective fusion algorithm. The algorithm can fuse different evaluation indexes and different test image databases, final result data obtained through the algorithm can reflect the comprehensive efficiency of various objective evaluation algorithms of the image quality so as to provide a basis for the screening and the application of various algorithms.

Description

Technical field [0001] The present invention relates to the computer field, in particular to an algorithm for evaluating the overall efficiency of an objective method of image quality. Background technique [0002] Images convey a large amount of information and play an important role in digital electronic products and Internet applications. Along with massive image data, we are faced with problems such as rapid screening of information, effective analysis of image content, and accurate judgment of image quality. The objective evaluation method of image quality uses a computer to simulate the human visual system to model and study image quality. This method has become the focus of research in this field with an automatic and continuous and efficient work method. [0003] At present, the objective evaluation algorithm of image quality has achieved considerable development. Traditional classic evaluation algorithms such as root mean square error RMSE, signal-to-noise ratio SNR and ...

AI Technical Summary

Problems solved by technology

In addition, for the same image quality objective evaluation algorithm, the evaluation result level is also quite different for different test image databases.

However, there are very few literatures that specialize in the evaluation of the performance of various algorithms. The existing research results are limited to the three accuracy indicators of Spearman rank order correlation coefficient SROCC, root mean square error RMSE and Pearson correlation coefficient PLCC. The resulting problems are: 1. The time performance is a very important index to evaluate the performance of the algorithm, but there is no research result on the time performance of the image quality objective evaluation algorithm, so it is impossible to compare the time performance of various evaluation algorithms ; 2. Stability is an important performance index of the algorithm, but there is no research result on the stability evaluation of the image quality objective evaluation algorithm; 3. The same image quality objective evaluation algorithm has different performance for the evaluation of different types of distortion images , how to design a simple and effective fusion algorithm to obtain a comprehensive evaluation of various distorted images, there is no research result; 4, the existing three accuracy index evaluation standards are not uniform, the correlation coefficient SROCC, and Pearson correlation The three value ranges of the coefficient PLCC are all 0 to 1, and the larger the value, the higher the accuracy, but the root mean square error RMSE value range is >1, and the larger the value, the lower the accuracy; 5. The same type The image quality objective evaluation algorithm has different performances for the evaluation of different test image databases. How to design a simple and effective fusion algorithm to obtain comprehensive evaluation of various test databases has no research results;

[0005] Taken together, there are very few research results on measuring the comprehensive efficiency of image quality objective evaluation algorithms. The reasonable selection and promotion of the algorithm also affects the research progress in the field of objective evaluation of image quality.

Images