SAR Image Recognition Method Based on Bayesian Network

Abstract

The invention discloses a SAR image recognition method based on Bayesian network. It mainly solves the problem that there are few types of target recognition and training samples are required in the prior art. The implementation steps are: 1. According to the method based on hierarchical visual semantics and adaptive neighborhood polynomial implicit model, the SAR image segmentation result is obtained; 2. According to the SAR image segmentation result, the corresponding relationship between the image gray level and the target category is constructed ; 3. Construct a Bayesian network according to the corresponding relationship between the image gray level and the target category; 4. Determine the target category of the segmented area according to the Bayesian network. The invention realizes good recognition effect of SAR images and can be used for target tracking.

Description

technical field [0001] The invention belongs to the technical field of image processing, in particular to a SAR image recognition method, which can be used for image fusion, registration and target tracking. Background technique [0002] SAR is an important means of earth observation, and it is of great significance to use SAR images for target recognition. Typical SAR image recognition methods are based on template matching. This type of method mainly extracts features from the target in the SAR image, such as shape features, boundary features, and grayscale features, and then establishes a template for the target. According to the template, a similarity measure is designed to identify the types of objects. This type of method has a strong dependence on the target template and is less flexible. When the target has a large deformation, it cannot recognize the target very well, and the type of target recognized by this type of method is relatively single. [0003] In orde...

AI Technical Summary

Problems solved by technology

However, for SAR images, it is quite difficult or even impossible to obtain a large number of training samples, thus limiting the application of this type of method in SAR image recognition

In addition, this type of method does not consider the causal relationship between the characteristics of the SAR image and the target category, so the target category recognized is relatively simple.

Images