Font Size: a A A

The Study Of Target Classification Method Based On Low-Resolution Radar Return Sequences Outline Image

Posted on:2005-10-04Degree:MasterType:Thesis
Country:ChinaCandidate:Q S LinFull Text:PDF
GTID:2168360155472000Subject:Information and Communication Engineering
Abstract/Summary:PDF Full Text Request
Along with the offshore situation becoming complicated day by day, the target recognition that simply relys on manual classification can not meet the need of surveillance in real time, and the target recognition that can carry out in the course of Track While Scan (TWS) becomes needed for the radar station in front line. The researching backdrop of the dissertation is actualizing the automatic target classification in the course of TWS, and the recognition method is studied that bases on the return sequences outline image of low-resolution radar. The dissertation is made up of three parties: the studying of the electromagnetism scattering in course of low-resolution radar wave bundle scanning target, targets extraction and features extraction, classification. First of all, based on the characteristic of electromagnetic scattering in optics, the electromagnetic scattering is researched when the target is scanned by low-resolution radar, and the theory is studied why the return sequences outline image is such. Then, on basis of the researches of electromagnetic scattering and the chatracteristic of return sequences outline, the target is detected and extracted, and the features that can reflect the target' s characteristic are extracted. Finally, according to the distribution characteristic of the targets in the characteristic space and the requisition of terseness and high availability to the recognition algorithm, the clustering method of C-Means is improved by the thought of layered clustering, and the classification system is carried out based the clustering method.Proved by the experimental results using IRIS data and real data, the method put forward in the dissertation is effective generally and effective for the target recognition in the course of TWS based on low-resolution radar return sequences outline image, and the excellence including high accuracy, real time, and concision is tested too.In the dissertation,the research on target electromagnetic scattering, target detection and extraction, classification method have universal significance and have certain reference value to other similar problem.
Keywords/Search Tags:Low-Resolution Radar, Outline Image, Target Recognition, Electromagnetism Scattering, Target Extraction, Feature Extraction, Clustering
PDF Full Text Request
Related items