SAR image segmentation method based on area division and self-adaptive polynomial implicit model

Abstract

The invention belongs to the technical field of image processing and provides an SAR image segmentation method based on area division and a self-adaptive polynomial implicit model. The method includes the following steps: 1. using an obtaining method of an initial sketch which is based on an SAR image and includes a multiplicative speckle noise model to extract an SAR image sketch; 2. according to an area chart, dividing the SAR image into a nonstructural area and a structural area of a pixel space; 3. according to semantic information based on rule-based reasoning, dividing the structural area of the pixel space into a boundary area and a non-boundary area; 4. establishing an image segmentation method of a multilayer polynomial implicit model on the non-structural area of the pixel space; 5. proposing an image segmentation method of a single-layer polynomial implicit model, which is based on a geometric structure window and based on a square window, on the structural area of the pixel space; 6. combining the segmentation results of the different areas and obtaining a needed segmentation result. The SAR image segmentation method based on area division and the self-adaptive polynomial implicit model realizes an excellent segmentation effect of a high-resolution SAR image and is applicable to segmentation of a high-resolution SAR image.

Description

technical field [0001] The invention belongs to the technical field of image processing and relates to a high-resolution SAR image segmentation method, which can be used for high-resolution SAR image segmentation, in particular to a SAR image segmentation method based on area division and an adaptive polynomial hidden model. Background technique [0002] Synthetic Aperture Radar (SAR) technology has developed rapidly in recent years, and the acquisition of massive high-resolution SAR images has brought challenges to the understanding and interpretation of SAR images. SAR image segmentation is a crucial step in SAR image understanding and interpretation, and lays the foundation for SAR image target recognition and tracking, which has important application significance. Due to the imaging characteristics of SAR images, SAR image segmentation mainly has the following difficulties. The first is the multiplicative noise of SAR images, which is an inherent characteristic of the S...

AI Technical Summary

Problems solved by technology

The polynomial implicit model is a relatively new method in the model-based method. In the model, both the magnitude feature and the texture feature of the SAR image are considered, and the finite mixture model is used to combine the two, and a good segmentation result is obtained. However, the space The relational model is based on the window. When the window is too large, the regional consistency of the segmentation result is better, but it will cause the loss of detail information, especially the loss of point objects and line objects; when the window is small, the detail information remains better. , but the regional consistency of the segmentation results will decrease

Hierarchical polynomial hidden model takes into account the multi-scale information of the image, and the transfer between layers can better capture the larger-scale content in the image, but it will inaccurately locate the image boundary and even cause the loss of point targets and line targets, which is not conducive to SAR image understanding and interpretation

The defect of these methods is that the structural information of the image is not considered, and the different structural regions of the image are segmented with the same strategy, and it is difficult to balance the regional consistency and the maintenance of detail information.

Images