Synthetic aperture radar image self-adaptive spot noise suppressing method

Abstract

The present invention proposes a synthetic aperture radar image adaptive speckle noise suppression method, and proposes to use a global statistical quantity heterogeneity entropy He as a criterion for judging regional homogeneity and heterogeneity, because the heterogeneity entropy He utilizes the image The global probability and statistics distribution can measure the changes of edge details and texture information very well, so as to judge the local homogeneity and heterogeneity of the image more accurately, and adaptively take different noise reduction measures. This method can It suppresses speckle noise in synthetic aperture radar images without sacrificing spatial resolution and preserves the texture details of the target edge.

Description

technical field [0001] The invention belongs to the field of image processing, and relates to a speckle noise suppression method, in particular to an adaptive speckle noise suppression method for synthetic aperture radar images Background technique [0002] Synthetic Aperture Radar (SAR) is not limited by factors such as weather, geography and time. It can perform high-resolution imaging on the ground and has a certain penetrating power. Therefore, it is widely used in military reconnaissance, resource detection, fire, ocean observation, environmental monitoring, etc. [0003] Since SAR works based on the coherent imaging mechanism, there is serious speckle noise in the SAR image, which has a great impact on the interpretation and information extraction of radar images. Early SAR image speckle noise suppression was achieved through multi-view processing. , the main disadvantage of this method is that the spatial resolution of the image is sacrificed, and it is not suitable f...

AI Technical Summary

Problems solved by technology

[0003] Since SAR works based on the coherent imaging mechanism, there is serious speckle noise in the SAR image, which has a great impact on the interpretation and information extraction of radar images. Early SAR image speckle noise suppression was achieved through multi-view processing. , the main disadvantage of this method is that the spatial resolution of the image is sacrificed, and it is not suitable for the current high-resolution SAR system

[0004] At present, the spatial filtering method is usually used to suppress speckle noise. The spatial filtering method generally uses a sliding window to weight the pixels in the window to obtain the pixel value of the center point of the window. Traditional methods, such as Lee filtering, Kuan Filtering, Frost filtering, and Gamma MAP filtering, but the images obtained by the above methods have problems such as blurred edges and texture details. Recent methods, such as the improved Lee filtering, use local statistics, that is, the ratio Cv of the mean and standard deviation under the sliding window as The weight factor, in the uniform area and the edge area of ​​the comparative mutation, outputs different filter values ​​to suppress the speckle noise. This improved Lee algorithm can maintain the edge of the mutation to some extent, but the speckle noise in the general SAR image not only contains Multiplicative noise also includes additive noise, and Cv is only suitable for multiplicative noise models, which is very sensitive to additive noise and is only a local statistic, so edge details and texture information are still not well maintained

Images