Region-Based Color Fusion Method for Visible and IR Image Sequences

A novel region-based color mapping method is proposed to render fused image of visible and infrared (IR). The method is based on image segmentation, region recognition, image fusion and color transfer. Firstly visible image and IR image are fused based on the curvelet transform. At the same time, a color database is formed by grouping a set of natural color images according to their scene contents. A false color image obtained by assigning the source images to RGB channels is segmented with the multiscale normalized cuts method. Then each segmented region is recognized by support vector machine (SVM) and automatically associated with a natural color image. Finally the colors of the natural color image are transferred to the segmented region in a perceptually decorrelated color space. Experiments show that the image colorized by the proposed method resembles the natural color appearance and will help the observer interpreting the scene more intuitively than produced by other coloring methods.