摘要
利用小波分析技术提取的特征,设计了SOM神经网络识别器,并用常规的BP神经网络识别器来比对分析,考察SOM神经网络识别器应用在声目标识别上的可行性。试验结果证实,SOM神经网络应用在声目标识别系统是可行的。
With the features extracted from wavelet analysis, SOM neural network identifier is designed and compared with the conventional BP neural network to test the feasibility of the SOM application in sonic target recognition. The results verify that the SOM neural networks can be used in sonic recognition system.
出处
《长春工业大学学报》
CAS
2008年第5期492-495,共4页
Journal of Changchun University of Technology
关键词
声目标识别
SOM神经网络
小波分析
sonic target recognition
SOM neural network
wavelet analysis.