The Method of Multi-attribute Target Recognition Based On Laser Detection
Jianbiao Chen, Huayan Sun, Yong-hui Zheng · Advances in intelligent systems research/Advances in Intelligent Systems Research · 2015
Ideal result of target recognition can't be obtained just with single attribute.In order to improve the performance of target recognition and reflect the physical attributes of the target, some attributes are extracted from the high range resolution profile of the unknown target to form an multi-attribute vector, such as the frequency amplitude, the length of energy gathering, the centre moments and the bispectrum and so on.Because of the high dimensions of the vector, the PCA (Principal Component Analysis) method is used to decrease the dimension.The processed vector is regarded as the input of the BP (Back Propagation) neural network to obtain the result of target recognition.The simulation result indicates that it is helpful for improving the target recognition rate with the multi-attribute extraction method and PCA method.