Radar high-resolution range profile feature extraction method based on multiple kernel projection subspace fusion

For radar automatic target recognition, the features should be extracted with sufficient target information, high discrimination, noise robustness and low feature vector dimension. In this study, a novel method for HRRP feature extraction is proposed, named as multiple kernel projection subspace fusion (MKPSF) method. For MKPSF, the statistical characteristics of both strong and weak scattering range cells are effectively exploited with an improved projection subspace extraction method. Therefore, the integrity of target information and the noise robustness can be ensured for HRRP recognition. In addition, a novel criterion function is constructed based on feature fusion and Fisher criterion theories. In this criterion function, the within-class correlation and between-class discrimination are maximised to guarantee high discrimination for feature vectors. Besides, the redundancy and dimensionality of the feature vectors are reduced by a fusion operation within the criterion function. This facilitates the reduction of the computational complexity in the practical radar target recognition system. Experimental results with measured datasets validate the efficiency of the proposed method.