Perceptual Image Fusion Technique of RGB and NIR Images

Image fusing techniques for visible-color (RGB) and near-infrared (NIR) images have been actively studied in last decades. Noting that the information obtained from the NIR images can improve the visibility of the RGB images, we propose an algorithm that fuses RGB image with NIR to improve the quality of RGB image based on perceptual information. We first decompose the RGB/NIR image into two parts, detail and base. Subsequently, we adjust each part by applying Just-noticeable difference (JND) profile computed by RGB image. The perceptual information quantified in terms of JND is utilized to preserve the original color characteristics by preventing saturation of image pixel values. Experimental results show that the proposed method provides better texture information and contrast adjustment than the conventional tone-mapped HDR images.