A low-illuminance image enhancement method in YUV color space

Abstract

Aiming at the problems of low-illuminance images such as low brightness, blurred details, and poor visual effects, a low-illuminance image enhancement algorithm based on YUV color space is proposed. First, convert the image from RGB space to YUV space, perform multi-scale guided filtering on the Y component to obtain the illuminance component, use homomorphic filtering to adjust the illuminance component, and then obtain the reflection component through Retinex theory; then use multi-scale bilateral filtering to obtain the detail image of the reflection component; finally, the detail image is added to the reflection component and adjusted by adaptive gamma correction to obtain an enhanced image. The results of the experiment subjectively obtained images with high brightness, clear details, and conform to human visual effects, and objectively performed well in terms of average brightness, information entropy, and average gradient. The results show that the algorithm can effectively increase image brightness, enhance image details, and improve image visual effects.