A new sense-through-foliage target recognition method based on hybrid differential evolution and self-adaptive particle swarm optimization-based support vector machine

Sense-through-foliage target detection and recognition is of interest to both military and civilian research. In this paper, a new recognition method based on hybrid differential evolution and self-adaptive particle swarm optimization-based support vector machine (SVM) is proposed to recognize targets obscured by foliage. To seek the optimal parameters of SVM, a new hybrid differential evolution and self-adaptive particle swarm optimization (DEPSO) algorithm is developed to determine the optimal parameters for SVM with the highest accuracy and generalization ability. In this work, sparse representation is applied to extract the target features from real target echo waveforms measured by a bistatic ultra-wideband (UWB) radar system. Then, the extracted features are input into the proposed method to automatically recognize the types of targets. This method is validated by experiments taken in the forest environment. Compared with the commonly used particle swarm optimization-optimized SVM (PSO-SVM), SVM, k-nearest neighbor (KNN) and BP neural network (BPNN), the proposed DEPSO-SVM can achieve a higher accuracy. Experimental results demonstrate the effectiveness and robustness of the proposed method for sense-through-foliage target recognition.