Reasearch On Target Detection Technology With Complex Background

With the use of advanced system radar and the increasing complexity of radar application scenarios,the detection background is becoming more and more complex,forming not only one of the uniform environment,the multi-target environment,or the clutter edge environment,but multiple environments coexisting.The detection environment based on uniform environment makes the CFAR detector with superior detection performance in a single environment unable to meet the target detection requirements of coexistence in multiple environments,which promotes the birth and development of intelligent CFAR detectors.In this paper,for the case of detecting the background distribution is not unique,the background is segmented based on KLD(Kullback–Leibler divergence),the distribution type of data in different regions is identified and unified into an exponential distribution.Then,the measured data processing method is used to improve the TM-CFAR detector,so that the number of deleted large values 2r in the TM-CFAR can be changed adaptively with the number of interference targets,without artificial setting.The specific work and research contents of this article are as follows:(1)Study the performance of classic CFAR detector-mean CFAR detectors: CA-CFAR,GO-CFAR,SO-CFAR and ordered CFAR detector OS-CFAR and mean CFAR and ordered CFAR detectors uniform framework: TM-CFAR,and the detection principles of S-CFAR and VI-CFAR are also given.(2)Use KLD to calculate the statistical distribution difference of the measured data,and then use the Ostu algorithm to calculate the automatic segmentation threshold,so as to divide the RD spectrum data into noise areas and clutter areas,respectively to identify and model statistical models,use MSE test and KS,and then converts the non-exponential distribution data into exponential distribution data.The statistics of the actual false alarm rate show that the converted data is better than the converted data.The former is more conducive to the classic CFAR detector to maintain a constant false alarm rate.(3)Based on KLD combined with Ostu method,on the basis of TM-CFAR,this paper improves the TM-CFAR detector to make it have the characteristics of adaptive CFAR detector.Simulation experiments show that the detector overcomes the problems of the tolerance of the number of interference targets and the number of samples removed cannot be automatically adjusted with the actual number of interference targets.At the same time,there is no requirement for the location of the interference target.In a multi-target environment with a large number of interference targets,it can maintain better detection performance than the classic CFAR detectors and S-CFAR and VI detectors.