Segmentation Method of Polarized SAR Image Based on Low Rank Decomposition and Histogram Statistics

Abstract

The invention discloses a polarized SAR image segmentation method based on low rank decomposition and histogram statistics, mainly solving the problem that the consistency is different in the present segmentation technology area. The realization steps of the polarized SAR image segmentation method comprises steps of successively extracting a power graph and a sketching graph of a polarized SAR image, utilizing the sketching graph to extract an area graph of the polarized SAR image power image, extracting a low rank observation array in an aggregated area corresponding to the polarized SAR image power image, performing low rank decomposition, performing histogram statistics on the low rank part, constructing a similarity array, utilizing a similarity array to segment the aggregated area, performing segmentation on a formal region and a structural region, and merging the partitioning results of the aggregated region, the formal region and the structural region to obtain the segmented polarized SAR image. The segmentation result of the invention has good regional uniformity, and the invention improves the segmentation result of the polarized SAR image and can be used for target detection and identification.

Description

technical field [0001] The invention belongs to the technical field of image processing and remote sensing, and relates to a method for segmenting polarimetric SAR images, which can be used for subsequent recognition of polarimetric synthetic aperture radar POLSAR images. Background technique [0002] The polarimetric synthetic aperture radar POLSAR image segmentation method refers to distinguishing different regions according to the gray level, polarization information, structure, aggregation and other characteristics of the image. The research on the image segmentation method of polarimetric synthetic aperture radar (POLSAR) has very practical significance, and has very important applications in military affairs and agriculture. The image segmentation of polarimetric synthetic aperture radar POLSAR is an important basis for polarimetric image processing and interpretation, and the quality of its segmentation will directly affect the subsequent analysis and identification w...

AI Technical Summary

Problems solved by technology

[0004] The above methods can make good use of the statistical characteristics and scattering mechanism of polarimetric SAR data for classification, but these methods are mainly pixel-based classification methods that do not take advantage of the semantic information of polarimetric SAR images and the low rank of semantic regions themselves. structure for semantic segmentation

Therefore, the above-mentioned traditional polarization SAR image segmentation method has many defects: (1) the regional consistency of the same object is not good, resulting in speckle noise; It is difficult to classify such ground objects into one category from the perspective of human vision and image understanding.

Images