In general, details of night images are easily blurred. In order to improve the quality of the night image, a new Retinex algorithm is proposed based on the YUV color space with chrominance components correction strategy. Firstly, the image is converted from RGB color space to YUV color space. Then the brightness component Y is enhanced by using multi-scale Retinex algorithm. At the same time, chrominance components U, V are corrected based on Y component after enhancement. Finally, the image is converted to RGB color space to output. Results show that while image details are enhanced the color has been maintained. By comparison with other Retinex algorithms, experimental results show that the algorithm applied to night image enhancement has achieved good results in image color fidelity and detail information enhancement respects.