基于图像特征融合与决策融合的多模式人脸识别方法英文.pdf

7 1 Vol.7 No.1 2009 1 Nanotechnology and Precision Engineering Jan. 2009 Multimodal Face Recognition Based on Images Fusion on Feature and Decision Levels LIU Jin ZHANG Le-shi XU Ke-xin State Key Laboratory of Precision Measuring Technology and Instruments, Tianjin University, Tianjin 300072, China Abstract Face recognition based on image fusion from visual and infrared images is a new study focus of the multimodal face recognition. In this paper the fusions of visual and infrared images on feature level and decision level were discussed. On the feature level the feature fusion was realized due to the effective dimension reduction using genetic algorithmGA on the decision level the fusion method based on Dempster-Shafer evidence theory was proposed. Totally 1 000 pictures of 50 subjects were taken for the fusion experiment, with 10 visual images and 10 infrared images for each person. The experi- mental results show that the feature fusion and the decision fusion can improve the correct recognition rate compared with single type image. The correct recognition rate using LDA and D_LDA reaches 100%. Therefore the feature fusion based on GA and the decision fusion based on Dempster-Shafer evidence theory are effective methods to realize multimodal face rec- ognition. Keywords multimodal face recognition image fusion genetic algorithm Dempster-Shafer evidence theory visual image infrared image