Research on the Algorithm of Object Recognition of Ship noise Based on Auditory Feature and Pattern-matching

In this paper, an object recognition algorithm of ship noise is presented combined a new kind of information processing technology with pattern matching. The procedure of processing acoustic signal of human being's hearing is imitated to extract the auditory correlation spectrum (ARS) as feature in the algorithm. Then, a novel feature named as summation of auditory correlation spectrum (SARS) is defined. Consequently the simulation on the presented algorithm and that based on power spectrum and based on neural network is carried out. The results illustrate that the given algorithm is effective and efficient