Low-Illumination Images Enhancement Using an Optimized Retinex Model with Improved Bilateral Filtering

Images taken in low light conditions suffer from poor quality, which prevents visual systems from fully fulfilling their purposes. To improve the quality of these images, this paper proposes a correctional method based on improved bilateral filtering and optimized gamma transform. Firstly, the original image is converted from RGB to the HSV space, and a modified bilateral filter function is used as the center surround function of Multi-Scale Retinex algorithm. After which, the image brightness component is decomposed to obtain the illuminated component and the reflected component. An optimized Gamma transform function is proposed to enhance the radiation component of the image. For the reflected components, the high-frequency noise is filtered by non-local mean filtering, and the details are enhanced by edge enhancement algorithms such as the Laplace operator. Finally, the adaptive correction formula adjusts the saturation component to ensure that the color of the image is uniform. Experimental results on several challenging low illumination image datasets show that the proposed algorithm is effective in terms of computation time, gray mean, information entropy, average gradient and peak signal-to-noise ratio.