In order to improve the enhancement effect of partial brighter and darker images, an improved gray image enhancement method is proposed, which is based on histogram equalization and single-scale Retinex (SSR) algorithm. Improved histogram equalization enhancement for original images is carried out, and the low frequency components of the original images are enhanced by histogram equalization and fused with the high frequency components. SSR algorithm enhancement is carried out to original images. Weighted fusion of the two images is carried out. Results show that the image obtained by the new method is more prominent to highlight details of the original image, and the visual effect is improved significantly. Contrast and entropy are enhanced, and the brightness is adjusted. The feasibility on medical X-ray images is verified. Compared with other enhancement algorithms, the new method is more conducive to the accurate analysis of images.