Research on image fusion for visible and infrared images

Abstract

According to study on the definition of information fusion and the analysis about the nature of image fusion, the image fusion method is divided into spatial domain and frequency domain. By analyzing and comparing the performance of the both two, the advantages and prospects of the latter are confirmed. In frequency domain, the fusion method based on two-dimensional discrete wavelet transform is mainly introduced. Based on high and low filters, it decomposes the source image into couples of sub-band images which highlight image's detail information in different spatial resolutions. By means of applying different fusion rules to distinct sub-band images, the final fused image is achieved by using the inverse discrete wavelet transform. The experimental results indicate that the fused image successfully reflects the comprehensive information on the target object or the target scene.