Coherent Speckle Suppression Method for Non-local Mean Radar Image Based on Ratio Distance

Abstract

The invention provides a non-local mean radar image coherence spot suppression method based on ratio distance, which includes: reading the synthetic aperture radar coherence spot image; setting the corresponding search window and similarity window size; according to any two similarities The gray value in the window determines the ratio distance of any two similarity windows; the gray value ratio of the center points of any two similarity windows is determined according to the gray value of the center point of any two similarity windows. ; Determine the spatial distance between the center points of any two similarity windows according to the spatial coordinates of the center points of any two similarity windows; according to the ratio distance, the gray value ratio of the center point and the center point The spatial distance calculates the weight value of the pixel point in the two similarity windows; the estimated value of the current pixel point is estimated through the weight value. The invention can effectively suppress coherent speckle noise, effectively maintain edges, improve performance indicators, and have better visibility.

Description

Technical field [0001] The embodiment of the present invention relates to the field of radar image coherent speckle suppression, and in particular to a non-local mean radar image coherent speckle suppression method based on ratio distance. Background technique [0002] In the traditional non-local mean algorithm, using the Euclidean distance of the gray value of the image block as the calculation similarity is no longer suitable for the multiplicative model. After rigorous theoretical derivation, it can be known that the ratio distance is more suitable for the multiplicative model and can maintain The true correlation between pixels. However, when calculating the similarity using the ratio distance, the closer the distance is to 1, the greater the similarity, and the farther away from 1 the smaller the similarity. Therefore, the ratio distance cannot be directly used as a measure of the similarity of two image blocks. [0003] Traditional coherent speckle suppression methods inclu...

AI Technical Summary

Problems solved by technology

[0004] The speckle suppression method in space domain is over-smoothed in the strong edge and texture area, resulting in reduced suppression performance; wavelet and anisotropic diffusion coherent speckle suppression will appear some interference blocks in the radar image, affecting the visibility; the current non-local mean method only The similarity is measured by the ratio distance, and the speckle suppression performance is not stable enough

Images