SAR image segmentation method combining background information and maximum posterior marginal probability standard

Abstract

The invention discloses an SAR image segmentation method combining background information and a maximum posterior marginal probability standard, belonging to the technical field of image processing, and mainly solving the problem that the image margins with small gray level difference can not be acquired completely and segmented accurately by the existing watershed segmentation method based on a regional hierarchical mode. The segmentation process comprises the following steps: carrying out mean sampling the original image to obtain a quadtree hierarchical structure; intercepting a training sample on the original image to carry out feature extraction; carrying out gradient corrected watershed segmentation on each other hierarchies except the coarsest hierarchy; combining the regions obtained by the watershed segmentation; regarding a parent region and an adjacent region as the background information; training parameters by using the maximum expected value; and calculating to obtain the posterior marginal probability P (xA|y) of the regions at each hierarchy, so as to obtain a final segmentation result. The method of the invention has the advantages of good regional consistency of the segmentation result and accurate margin, and can be used in SAR image target identification.

Description

technical field [0001] The invention belongs to the field of image processing, and is an image segmentation method, which is suitable for synthetic aperture radar image SAR and can be applied to target recognition. Background technique [0002] Image segmentation has always been a major difficulty in image understanding and interpretation. Segmentation is an important step from image processing to image analysis, and is the basis of object classification and recognition. Essentially, the SAR image reflects the electromagnetic scattering characteristics and structural characteristics of the target. The particularity of SAR imaging makes the segmentation method for this type of image different from that of ordinary optical images: SAR images contain a lot of coherent speckle noise, while conventional segmentation methods are usually highly sensitive to noise and are not suitable for such images; The electromagnetic wave scattering characteristics of the remote sensing area to...

AI Technical Summary

Problems solved by technology

[0005] 1) The optimal threshold segmentation method based on the histogram, because the SAR image is a non-optical imaging method, the gray level of this method changes slowly, so this method often has little effect

[0006] 2) The SAR image segmentation method based on edge detection. Due to the multiplicative coherent speckle noise in the SAR image, this method makes the gray value of the edge pixels fluctuate greatly, and it is difficult to accurately locate the edge pixel position.

[0007] 3) The SAR image segmentation method based on regional block merging, this method is not good in statistical separation of segmented regions, and it is easy to cause over-segmentation

[0010] 1) The segmentation method based on the combination optimization model. After establishing the combination optimization model of SAR image segmentation, this method solves the global optimal solution for the objective function. However, since the objective function is usually a multivariate function, it is very easy to fall into the local Optimal solution

[0011] 2) The segmentation method of the pixel-based Markov random field model. Due to the difficulty in estimating the parameters of the Markov random field, this method requires iteration, the calculation speed is slow, and the statistical relationship between different resolutions of SAR images is not fully explored. Only a single The prior probability problem in the case of resolution, the segmentation accuracy is low

This method takes the area as the processing unit, and uses the multi-scale model to mine the statistical correlation characteristics of the same observation scene in the SAR image at different resolutions. It is independent and does not make full use of the information of adjacent areas in the layer, which will lead to unreliable information transmission between layers and a high rate of area misclassification

Images