Radar Target Detection Based On Online Extraction Of Background Information

This thesis is under the background of the airborne forward-looking radar system, aiming at practical engineering. It mainly carries on the analysis of the radar detection background to design more suitable radar target detection method in complex clutter environment. Compared with ground-based radar, airborne radar has larger vision, more maneuverability and farther detection distance of low-altitude slow-flying target. However, due to the look down mode of airborne radar, its echo has a strong ground clutter. Bedises, the movements of machine and the changes of antenna scanning range cause the background detection to be more complicated. The detection background of airborne radar contains homogeneous and non-homogenous regions, which lead to the coexisting of many kinds of complex clutter distributions in the detection background. And these bring great difficulties to the airborne radar target detection.Therefore, this thesis analyses the characteristics of airborne radar detection background and the difficulties of target detection in detail. According to the complexity and diversity of airborne radar detection background, this thesis proposes an airborne radar target detection system based on the online extraction of background information, in order to solve the problems of lower detection probability and higher false-alarm probability on the background of airborne forward-looking radar complex clutter. This system includes the analysis of clutter characteristics and target characteristics of airborne radar detection background, the segmentation of detection background region, the online extraction of background information and multi-strategy CFAR detection. The specific contents of this thesis are as follows.1.This thesis studies airborne radar clutter characteristics and target characteristics, introduces the complex features of detection background, analyses the mechanism of uniform regions and non-uniform regions, and also analyses the broadening of target in the distance spectrum and Doppler spectrum under the background of the measured data. All of these lay the foundation for subsequent online extraction of background information and target detection.2.Two methods of airborne radar detection background segmentation are introduced in this thesis. One is based on the homogeneous/non-homogeneous distribution and statistical distribution differences. And the other is based on the characteristics of the clutter images. The former can effectively measure the statistical distribution differences in different regions, and divide the background into uniform and non-uniform regions. The latter uses maximizing the separability of the resultant classes methods, and can adaptively determine the segmentation threshold of uniform and non-uniform regions. By segmenting the detection background, it is beneficial for the statistical analysis of the distribution of background clutter.3.This thesis also introduces the online information extraction of airborne radar detection background, and analyses the statistical characteristics of the regional detection background, according to the results of the detection background segmentation, which realizes the identification of the statistical model of detection background(clutter) and the estimation of statistical parameter. And then according to the statistical results, difference indexed of background detection data are normalized, which realizes the homogenization of detection background.4.According to the results of background information online extraction, this thesis proposes a background information-based adaptive CFAR(BIA-CFAR) detection method. This method can make full use of background information to adjust the detector design, in order to improve the detection performance of airborne radar. And through simulation experiments and test data analysis, this thesis compares radar target detection method which is based on online extraction of background information with traditional radar target detection method, and proves that the detection performance is improved.This thesis proposes the radar target detection method based on online extraction of background information. Through the online real-time extraction of detection background information, it improves the capability of radar detector to adapt to the complex background environment of target, and forms adaptive multi-strategy CFAR detection system to improve the detection performance of airborne radar target.