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

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 samp