超宽带探地雷达被广泛用于检测浅层埋藏的反人员地雷,其中滤除强的地表反射回波是完成目标检测、成像与识别的关键。将独立分量分析算法与数据非一致性检测方法相结合,通过分析独立分量的频谱和倒谱,给出了一种可以适应多目标数据区域且自动选取独立分量的地杂波抑制方法。实验结果证明了该方法的可行性和有效性。
The ultra wideband ground penetrating radar(GPR) is a useful tool for anti personnel mines detection, removing the ground cluttes with much more energy than useful signal is very important to following detection,imaging and target recognition. This paper presents an adaptive independent components(ICs) selection ground clutter suppression method for muhinomial targets data based on analysis of spectrum and cepstrum,which uses independent component analysis(ICA) algorithm combining with non-homogeneous detections(NHD). Experimental results based on field-test data show that the proposed method is feasibleity and effectivity.