Abstract: To address the lack of prior knowledge in the field of image enhancement, a novel prior knowledge based on the perceptual characteristics of the human visual system—referred to as the Narrow Dynamic Prior—is proposed. Furthermore, an improved histogram equalization algorithm guided by the narrow dynamic prior is introduced. Both subjective and objective experimental analyses were conducted on different grayscale transformation results. The applicability of histogram equalization and linear stretching to different types of images was analyzed, revealing that when an image exhibits narrow dynamic characteristics, the human visual system tends to prefer the linear stretching enhancement results. Based on this visual perception characteristic, narrow dynamic correction and gamma correction were applied to the image histogram to expand the applicability and control the enhancement intensity. Experimental results on seven publicly available datasets demonstrate that the proposed algorithm is suitable for various types of images, including low-light and blurred images, and its enhancement results achieve superior subjective visual quality. The proposed method outperforms 12 state-of-the-art traditional and learning-based approaches, with average Perceptual Index (PI) and Perception-based Image Quality Evaluator (PIQE) scores of 3.736 0 and 35.965 9, respectively. These results further validate the rationality of the narrow dynamic prior.