HRRP multi‐target recognition in a beam using prior‐independent DBSCAN clustering algorithm

Abstract

During the operation of monopulse radar, the high-resolution range profiles (HRRPs) of multiple targets may be overlapped, which will reduce the target recognition performance. In this study, a novel multi-target recognition method is proposed based on prior-independent density-based spatial clustering of applications with noise (PI-DBSCAN) algorithm. In the training phase, various features of training samples are extracted, among which the distribution of strong range cells is utilised to obtain the parameters of PI-DBSCAN algorithm while the others are used to train the support vector machine (SVM) classifier. In the test phase, PI-DBSCAN algorithm is exploited at the radial distance-azimuth plane to segment the test multi-target HRRP. Afterward, the features of the segmented HRRPs are extracted and fed to the SVM classifier to be recognised. The proposed method has no constraint on the target motion and does not need to set any parameters manually, which benefits its application. The experiment results of real measured data show that the proposed method is robust against noise and the increasing target HRRP overlap ratio comparing with traditional methods.