关键词: 图像融合, Curvelet变换, 亮度重映射, 红外图像, 可见光彩色图像

Abstract: A novel fusion algorithm based on the second-generation Curvelet transform for infrared image and color visible image is proposed.In this paper，color visible image is converted from RGB color space to IV1V2 and its intensity component is extrated，as well as apply Curvelet transform to the intensity component and infrared image.In the progress，luminance remapping techniques is used to remap curvelet low-frequency coefficient before they are fused in weighted average rule，and picking large value rule is used to fuse high-frequency coefficients，and Curvelet inverse transform is used to fusion coefficients obtaining fused Intensity component.Furthermore，the color space inverse convert is used to obtain fused images.The experiment result shows that the proposed algorithm can overcome the problem in traditional image fusion algorithm，where one source image has higher intensity which may cover the other background texture and detail features.And the algorithm can improve fusion of infrared image spectral information，also has more comprehensive evaluation and better visual effects.

Key words: image fusion, Curvelet transform, luminance remapping, infrared image, color visible image