Infrared imagery is highly valued for its wide range of applications in defense, scientific research, production, medicine and health, security and other fields. Infrared technology is able to work effectively in environments where visible light is impenetrable or illuminated insufficiently, which makes it excellent for night vision, temperature monitoring, target identification, medical diagnosis, etc. However, in the process of acquiring infrared images, infrared detectors are inevitably affected by a variety of interference noise, which not only reduces the image quality, but also affects the effectiveness of the subsequent image processing process. The cold reflection phenomenon in infrared imaging systems, the narcissus effect, is a kind of interference noise generated by the detector itself emitted by the infrared radiation reflected back to the detector. This phenomenon often leads to artifacts and blurring in the image, which seriously affects the visibility of the image, and thus the subsequent processing applications such as target detection, identification and tracking. This not only degrades the image quality, but also restricts the miniaturization and lightweighting of the system, especially in modern precision systems, and solving the cold reflection problem becomes the key to achieving high-performance infrared imaging systems.In order to solve the problems caused by cold reflection, this study aims to develop an effective image restoration method to eliminate the cold reflection effect on infrared images, so as to improve the quality of infrared images and the overall performance of the system. The problem of cold reflection is not only important in theory, it also affects the design and application of various infrared imaging systems in practice. As the application of infrared technology in national defense and industry continues to expand, the development of a method that can effectively suppress or eliminate the cold reflection noise is of great practical significance to enhance the performance of infrared systems, reduce the computational burden, and promote the progress of infrared imaging technology. Aiming at the cold reflection phenomenon, this study designs a mid-wave cooling infrared computational imaging optical system based on the system design principle and optimized evaluation function. The mid-wave cooling infrared system has become a key component of modern infrared imaging technology due to its superior performance in detection sensitivity and imaging quality. The designed optical system is modeled in detail using TracePro software to obtain radiance outgrowth data for the system. These data include the radiative properties, reflection coefficients, and other important parameters of the different optical components of the system that may affect the cold reflection phenomenon. After completing the TracePro modeling, the radiance data are imported into Matlab software and matched with the gray values of the simulated image to generate a high-precision simulated image that can reflect the cold reflection effect. Considering the difficulty of eliminating the cold reflection in the image restoration process, this paper proposes a Butterworth cross-section-based method. The Butterworth filter is widely used in the field of signal processing due to its smooth transition and less oscillation. In this study, the Butterworth cross-section method is applied to image processing to eliminate the image blurring caused by the cold reflection effect. The image incompleteness caused by cold reflection noise is effectively suppressed by this method, and thus smooth image recovery is achieved.In order to verify the effectiveness of the proposed method, a detailed simulation of the cold reflection phenomenon in simulated infrared images and real infrared images is carried out in this paper. In the experiments, infrared images under different scenarios are generated and the possible cold reflection effects in them are simulated. By comparing the degraded images with the images recovered based on the Butterworth cross-section method, it is found that the method can effectively reduce the cold reflection noise and significantly improve the clarity and visibility of the images (Fig.6-Fig.7). In addition, the image quality before and after the restoration was quantitatively analyzed from various perspectives such as image standard deviation (STD), peak signal-to-noise ratio (PSNR) and structural similarity index (SSIM). The experimental results show that PSNR and SSIM, on the other hand, are significantly improved after image restoration using the Butterworth cross-section method (Tab.3-Tab.4). These results fully demonstrate the effectiveness of the method in improving IR image quality and signal-to-noise ratio.The experimental results show that the use of the Butterworth cross-section method to remove cold reflection noise not only significantly reduces the computational workload, but also significantly improves the quality and signal-to-noise ratio of infrared images. This study provides a new technical path for solving the cold reflection problem and provides a valuable reference for the design of future infrared imaging systems. The Butterworth cross-section method proposed in this study has a wide range of applications in both theory and practice. Particularly in the military, medical and security fields where high-precision imaging is required, the method can significantly improve the imaging quality of the system, reduce unnecessary computational overhead, and help to realize the miniaturization and lightweighting of the system. In addition, the method can be further applied to other types of image processing tasks, such as noise suppression in remote sensing images and image enhancement in industrial inspection.