Performance Evaluation of Infrared and Visible Image Fusion Algorithms for Face Recognition

Fusion of infrared and visible image is a potential solution to improve face recognition performance. In this paper, we propose a new fusion method based on singular value decomposition (SVD) and apply it to multiple spectrum face recognition. The performance of the proposed SVD-based fusion method is compared with that of average fusion, Laplacian Pyramid Decomposition (LPD) based fusion and Discrete Wavelet Transform (DWT) based fusion methods. The performances of the four image fusion methods for face recognition are analyzed by statistical experiments and the results show that the SVD-based fusion method is better in most conditions.