In order to improve the performance of image enhancement under different illumination conditions, an effective image enhancement method was proposed, which was based on the improved histogram equalization and Retinex algorithm. First, the low frequency components of the image were obtained using the ideal low-pass filter and its dynamic range was optimized by the improved histogram equalization. Second, the high frequency components of the image were estimated by the guided image filtering, which replaces the Gaussian filter in Retinex algorithm. Finally, the estimated results were amplified by the linear processing method. Experimental results show that the performance of the proposed method was better than the traditional Retinex algorithm and improved histogram equalization algorithm by the subjective and objective evaluation. Furthermore, the proposed method effectively improves the image’s visual effects and intelligibility with strong robustness.