Infrared image enhancement technology based on human visual characteristics and adaptive side window bilateral filter

This is the authors' abstract. We don't add key points for this paper.

At present, the output image of infrared imaging system is high dynamic range image, which is difficult to display on traditional low dynamic range display device. And infrared images have the problems of concentrated gray distribution, small detail contrast and low contrast. In order to solve these problems and provide more details which make infrared images easy to observe, it is necessary to compress the dynamic range of infrared images and enhance the differences between details. In this paper, an adaptive side window bilateral filtering algorithm is studied. On the basis of edge preservation, the detail extraction performance and operation speed of the studied algorithm are improved. And an infrared image enhancement algorithm is constructed by combining the edge preserving filter with the salient characteristics of human vision. In this algorithm, adaptive side window bilateral filtering is used to separate the original image into base layer image and detail layer image firstly. In the base layer image, the adaptive dynamic range compression method based on Sigmoid function curve is adopted to improve the poor visual effect of processing results in some existing algorithms. In the detail layer image, visual saliency map is introduced to enhance the details and suppress noise amplification. The final result is obtained by weighted fusion of the processed base layer and detail layer. The simulation test of real scene infrared image have shown that the proposed algorithm in this paper can effectively enhance the infrared image, facilitate the visual perception of human eyes and has good robustness and stability.