TY - GEN
T1 - Ground moving target identification based on neural network
AU - Ming, Li
AU - Yuyan, An
AU - Chunlan, Jiang
AU - Zaicheng, Wang
PY - 2011
Y1 - 2011
N2 - With the development of science and chip technology, more and more attention is taken on more accurate and more intelligent recognition of the complex targets. Target identification is studied based on neural network method in this paper. Firstly, Wavelet analysis method is used for target feature extraction. 4 layers of wavelet decomposition and reconstruction are done for multiple signals, several groups of feature vectors have been obtained and they constitute the neural network learning sample set. Secondly, by analyzing and comparing a variety of BP algorithm, the resilient BP method is finally selected. Only 16 steps of training are needed to meet the error requirement by the resilient BP learning algorithm. Then, Bp neural network is designed and trained according to the signal characteristics. Finally, a recognition test is carried out. The test results show the recognition rate of 90% for the vehicles and 80% for the personnel.
AB - With the development of science and chip technology, more and more attention is taken on more accurate and more intelligent recognition of the complex targets. Target identification is studied based on neural network method in this paper. Firstly, Wavelet analysis method is used for target feature extraction. 4 layers of wavelet decomposition and reconstruction are done for multiple signals, several groups of feature vectors have been obtained and they constitute the neural network learning sample set. Secondly, by analyzing and comparing a variety of BP algorithm, the resilient BP method is finally selected. Only 16 steps of training are needed to meet the error requirement by the resilient BP learning algorithm. Then, Bp neural network is designed and trained according to the signal characteristics. Finally, a recognition test is carried out. The test results show the recognition rate of 90% for the vehicles and 80% for the personnel.
KW - neural networks
KW - personnel
KW - seismic signals
KW - vehicles
KW - wavelet analysis
UR - http://www.scopus.com/inward/record.url?scp=81355133042&partnerID=8YFLogxK
U2 - 10.1109/ICICIS.2011.114
DO - 10.1109/ICICIS.2011.114
M3 - Conference contribution
AN - SCOPUS:81355133042
SN - 9780769545394
T3 - Proceedings - 2011 International Conference on Internet Computing and Information Services, ICICIS 2011
SP - 439
EP - 442
BT - Proceedings - 2011 International Conference on Internet Computing and Information Services, ICICIS 2011
T2 - 2011 International Conference on Internet Computing and Information Services, ICICIS 2011
Y2 - 17 September 2011 through 18 September 2011
ER -