Low Light Image Enhancement Based on Luminance map and Haze Removal Model

In this paper, a novel and effective algorithm is proposed for noise reduction and contrast enhancement in low light images based on luminance map and haze removal model. The proposed method is divided into two steps: i) A combined denoising method using the improved guided filtering based on gradient information and median filtering is proposed to obtain the initial denoised image. ii)Considering that an inverted low light image presents quite similar to a haze image, the haze removal model is used to enhance the denoised low light image. The luminance component L is extracted to obtain the transmission map with the adaptive weight from the inverted denoised image which is applied to Lab color space. Then the classical quad-tree subdivision is utilized to estimate the atmospheric light, and then the de-hazed image is recovered by the haze removal model. At last, we can get the final enhanced image by inverting the de-hazed image back. The experimental results show that the proposed algorithm reduces the noise and enhances the contrast of the low light image more effectively and robustly than the conventional and the state-of-the-art algorithms.