Correction of overexposure utilizing haze removal model and image fusion technique

Abstract

This paper presents an efficient method for overexposure correction utilizing haze removal model and image fusion technique, which draws on the experience of HDR technique. Assuming an OE image can be modeled as a normal exposure image added up with a layer of asymmetrical colorful haze, its submerged information in OE regions is enhanced by an improved haze removal model based on dark channel prior. The enhancement result possesses better visualization in OE regions and color distortion to a certain extent. With the image fusion technique based on weighted least squares filters and global contrast-based saliency, the texture obtained in OE regions is utilized to restore the overexposure. The advantages of the selected image fusion technique are validated in the paper. In the experiments, the proposed method is compared with conventional methods to corroborate the performance. Both the subjective visualization and quantitative indicators show that the result is effective in correcting the overexposure without increasing pseudo-information and oversaturation.