Feature Extraction Method of Ultrasonic Signal Based on Wavelet Coefficients Cluster

Support Vector Machine is a very good solution to the classification problem of small sample, but when the input feature vector dimension is larger, the classifier has complex structure, long training time and degraded performance. In order to solve this problem, a feature extraction method based on wavelet coefficients cluster was put forward. All the wavelet coefficients was clustered, the energy value of wavelet coefficients in each cluster was calculated and used as the input feature vector of a classifier. The dimension of input data was reduced greatly, and at the same time the specific problem information was retained. Support Vector Machine was used to identify the defects in steel plate, the experiment results showed that the method has higher classification accuracy.