Image fusion of palmprint and palm vein: Multispectral palm image fusion

Abstract

The rattlesnake combines visible signals and infrared signals by instinct. This process, and the development of multisensor computer vision systems, has promoted the research into image fusion. In this paper we propose to use two methods for fusion of the images from multi-sensor imaging system with the objective of establishing some preprocessing algorithms in palm authentication system. The methods derive from opponent-processing and dual-tree complex wavelet transform (DTCWT). Opponent-processing is based on biological model, while dual-tree complex wavelet transform is based on mathematical model. The dual-sensor fusion system combines image from a visible light CCD camera with near-infrared image. Experiments on different fused images show that our methods are fast enough to be applied in real-time system. Mutual information of fused images and source images show that these two methods are more effective for image fusion than discrete wavelet transform (DWT) and shift invariant discrete wavelet transform (SIDWT).