SAR image non-reference quality evaluation method based on multi-view amplitude statistics characteristics

Abstract

The invention discloses an SAR image non-reference quality evaluation method based on multi-view amplitude statistics characteristics. The problems of high implementation difficulty and low stability in the prior art are solved. The method comprises the steps that 1 a training sample and a test sample are extracted from an SAR image database; 2 the global equivalent visual acuity, the local maximum equivalent visual acuity and the ratio of the global equivalent visual acuity to the local maximum equivalent visual acuity of SAR images are calculated; 3 the fuzzy correlation coefficient of the SAR images is calculated; 4 fitting square root gamma distribution in the homogeneity region in the SAR images is selected; 5 the eigenvectors of all samples are extracted; 6 an eigenvector threshold is set to classify noise images; 7 a quality evaluation prediction model is trained for SAR images with different noise pollution types; 8 the quality value of the test sample is calculated; and 9 the quality of the test sample is determined according to the test sample quality value. According to the invention, the practicability and accuracy of SAR image quality evaluation are greatly improved, and the method can be used for screening SAR images.

Description

technical field [0001] The invention belongs to the field of image processing, and particularly relates to a method for evaluating the quality of SAR images, which can be used to identify SAR image data with uneven quality levels, provide help for users to use SAR images, and provide a basis for the design of SAR systems. Provide feedback for improving the design. technical background [0002] Synthetic Aperture Radar (SAR) is an active microwave remote sensing imaging radar. Compared with optical images and infrared images, SAR is not affected by lighting, weather and other conditions, and can achieve all-day, all-weather, high-resolution, large-area imaging. Nowadays, the application of SAR is more and more extensive, and the development is more and more rapid. As one of the research hotspots in the field of radar remote sensing, SAR image interpretation has always been difficult, which affects the application of SAR images to a certain extent. SAR image quality is one ...

AI Technical Summary

Problems solved by technology

[0005] First, many algorithms are greatly affected by noise and have poor stability;

[0006] Second, many algorithms require specific test scenarios, specific SAR image product formats, or additional SAR system parameters. These requirements limit the application of the algorithm, increase the difficulty of algorithm implementation, and limit the practical application of SAR image quality evaluation. application

Images