Research of image enhancement algorithm based on multi-scale retinex

Abstract

Parking supplementary system has been such broadly used in some vehicle that we could drive and park more safely and conveniently. As light intensity has important effect on images from cameras which are mounted on vehicle’s four different directions in order to obtain panoramic view. To eliminate the influence of light to the image and obtain consistent brightness of image, this paper proposed mix-algorithm combined histogram equalization (HE) with multi-scale retina (MSR). This method makes equalization processed image to MSR, and gets a enhanced result at last. This algorithm can decrease effectively noise and enhance the brightness of image. The performance of output image is shorter time, smaller mean square error and higher peak signal to noise ratio than MSR. The experimental results demonstrate that the proposed method can not only counterpoise illumination but also meet real-time requirements. So the proposed mix-algorithm can apply in lack of brightness and higher real-time such as parking supplementary system, traffic monitoring system and security systems etc.