Wavelet Network On The Submarine Target Recognition

The submarine target recognition which is one of the hot and difficult questions in the current research has the great significance in the national defense. In recent fifty years, the technology about detection mines has developed fast. The anti-mine question is urgent and serious day by day. In the world, Anti-mine equipment development falls behind the sea mine weapon development. Hence, the development of mine detection is necessary and exigent.As a result of environmental factor influence, the noise is always existent in the echo signals. This dissertation has used the pretreatment method of wavelet package to filtrate out the interferential signals based on wavelet theory, then, extract the features from the echo signals. The process of feature extraction is to transform the echo signal to different feature space and extract the feature vectors which reflect the category of the input sample. The extracted features are the input modes to the classifier. This dissertation has used the methods of feature extraction based on wavelet theory.Wavelet transform has good characteristic in time-domain and frequency-domain. The neural network has functions of self-study and ability of automatically back-check. Wavelet network preferably combined wavelet transform and neural network into a unit. This dissertation has used wavelet network to recognize the target and validated it by the data gained in experimentation.