Abstract: To address the common issues of low contrast and blurred details in infrared images, an infrared image enhancement algorithm combining atmospheric scattering model and intuitionistic fuzzy sets is proposed. The algorithm estimates the transmittance using a multi-scale dark channel prior, optimizes the atmospheric light estimation by integrating a superpixel segmentation algorithm, introduces heat source identification for transmittance compensation, and employs guided filtering and Laplacian refinement to achieve edge preservation of the transmittance. Subsequently, a hierarchical enhancement framework is constructed; it utilizes intuitionistic fuzzy sets to enhance the base layer, combines the contrast-limited adaptive histogram equalization algorithm to improve the overall contrast of the infrared image, and fuses the enhanced base layer with the detail layer to obtain an infrared image with clear details and rich textures. Experimental results show that, compared with other existing methods, the algorithm improves the information entropy and average gradient by more than 0.38% and 6.42% respectively, and can effectively enhance the detailed information of infrared images and improve the global contrast.