Target Recognition Algorithm Based on Support Vector Machine of Optimum Parameters

Target recognition algorithm based on support vector machine of optimum parameters is put forward in this paper. Firstly, local surrounding-line integral bispectrum feature is extracted from the bispectrum of range profile of target. Secondly, parameter scope is obtained through experiment method, optimum parameters of support vector machine are gotten using genetic algorithm. Finally, support vector machine of optimum parameters is used to recognize three kinds of tanks. High recognition rate of this algorithm is gained through simulation experiment, even though the SNR is very low.