Low Intensity RGB Texture Enhancement Based on Near Infrared Image Using Perceptual Information

The texture information in the near-infrared (NIR) image is known to be effective in improving the visibility of visible-range image, thus, texture enhancement using NIR images has been investigated for the low contrast area. Since image contrast enhancement ultimately targets for human, many methods have been developed considering the Human Visual System (HVS)'s responses to luminance. In a low brightness region, the Just-Noticeable-Difference (JND), which statistically reflects sensitivity characteristics of human eye, is high, thus noise is relatively hard to notice. However, after the contrast enhancement, the noise is well recognized because the JND value decreases. In this paper, JND with respect to the background brightness is computed to properly address the characteristic of the image. We propose an image contrast enhancement method that adds texture components of NIR images to RGB image with using JND value obtained from RGB images as weights.