Improvement of Low-Light Images without Loss of Naturalness based on the Retinex Theory

Abstract

In modern era, images enhancement is paying very significant role in image analysis and synthesis. We have used the Retinex theory to remove the dark from the first image help improve the clarity of dim or hazy photos. After that, the picture haze must be eliminated, first inverted the image and applied the optimized de-haze on it. By image fusion of both the obtained images through Principal Component Analysis (PCA), a better-quality image was obtained from what we can see in the simulations, things have definitely gotten better.