摘要
A feature extraction method of underwater echo is proposed, which uses the redundancy property of SWT （the Stationary Wavelet Transform） and the steady property of SVD （Singular Value Decomposition）. Since using singular values of SWT coefficients matrix as feature vectors, with the K-L transform to multi-subspace signal obtained from SWT, it is a feature compression method essentially. In contrast to the method of sub-bands energy feature based on discrete orthogonal wavelet transform （DWT）, better results of lake trial data are acquired with this method： （1） Under the same sample and distance within-class, distance between-class is larger than former; （2） Correct recognition rates are also higher than former, whether the training and testing samples are chosen from the same lake trail or not; （3） Sample sets varying, the variation range of correct recognition rates is far less than the former. Thus this method can obtain more robust, effective features and better correct recognition results.
A feature extraction method of underwater echo is proposed, which uses the redundancy property of SWT （the Stationary Wavelet Transform） and the steady property of SVD （Singular Value Decomposition）. Since using singular values of SWT coefficients matrix as feature vectors, with the K-L transform to multi-subspace signal obtained from SWT, it is a feature compression method essentially. In contrast to the method of sub-bands energy feature based on discrete orthogonal wavelet transform （DWT）, better results of lake trial data are acquired with this method： （1） Under the same sample and distance within-class, distance between-class is larger than former; （2） Correct recognition rates are also higher than former, whether the training and testing samples are chosen from the same lake trail or not; （3） Sample sets varying, the variation range of correct recognition rates is far less than the former. Thus this method can obtain more robust, effective features and better correct recognition results.
