Method for detecting SAR image cluster target based on target number posteriority

Abstract

The invention belongs to radar remote sensing or image processing technologies, and particularly relates to a method for detecting an SAR image cluster target based on target number posteriority. Themethod comprises the steps of extracting interesting regions from an SAR image and extracting a strong scattering center in each interest region to estimate the range of possible target number; selecting an estimated value of the target number in the range and dividing the scattering center in each interesting region into target regions by using a segmentation algorithm based on the constraint ofthe number; analyzing the similarity of the target regions through the obtained spatial size and intensity distribution of the target regions, and calculating the posterior probability of the estimated value of the target number; finally, finding the estimated value with the largest posterior probability by iteratively comparing detection results under different target numbers, and using the corresponding target region as the cluster target detection result of the method. The method for detecting the SAR image cluster target based on the target number posteriority achieves the detection and segmentation of the cluster target in the SAR image, and can more accurately distinguish different individuals in the target cluster.

Description

[0001] technical field [0002] The invention belongs to the technical field of radar remote sensing or image processing, and in particular relates to a SAR image cluster target detection method based on target quantity posteriori. Background technique [0003] Traditional SAR image target detection methods are used to extract isolated target areas. When there are target clusters in the SAR image scene, this type of detection method cannot correctly extract each target. Therefore, it is of great significance to improve the detection method of the cluster target to improve the signal detection ability of the radar. With the advancement of SAR imaging technology and the increasing application of SAR image detection technology, the algorithm research of target detection in SAR images with more complex scenes has become a hot spot in recent years. A target in a SAR image is generally composed of multiple scattering centers. While extracting the scattering centers of the target, t...

AI Technical Summary

Problems solved by technology

The model extracted by this method is derived from the real image, which makes full use of the target information, but its shortcomings are also very obvious.

This type of detection method relies on the support of a large amount of training data, and it is difficult for SAR images to provide enough training samples

In addition, the target detection frame in the training sample needs to be suitable for the target detection of the test image, and when the source of the image used for training and testing is different from the SAR system, the posterior probability of the target detection frame may not be accurate

Images