Coherent LADAR range image for target recognition using Back-propagation neural network
Liu Zheng-jun, Qi Li, Qi Wang · 2010
Range, intensity and Doppler images can be obtained by using coherent LADAR (laser radar). Range image is closely relative to physical structure property of target surface, and it can reflect essential characteristics of targets. At present the research on automatic target recognition technology for LADAR is attention-getting. The Back-propagation neural network (BPNN) has been used in multi-layered feed-forward adaptive neural networks, deserving special attention. In this paper, BPNN is introduced to recognize Vehicles on the ground with different carrier to noise ratio (CNR) of clutters, different orientations of target, and different distances between the object and sensor. The results of simulation show that the BPNN is feasible and effective for the target recognition.