Research on infrared/visual image fusion based on HSV space adaptation

The current infrared/visible image fusion methods cannot adapt to different low brightness environments, and solve the problems of low contrast, weak detail information and poor universality. An adaptive image fusion and enhancement method based on HSV (Hue Saturation Value) space is proposed. Based on the HSV color space of the image, the method can better adapt to the image fusion of low brightness images under different environmental conditions and suppress the loss of image details by enhancing the main structure and edge details of low light image with adaptive image original conditions. Firstly, the infrared/visual images are input for feature extraction. Then the fusion image is processed and fused by Resblock, and the fused image is converted to HSV space for HSV adaptive processing. Finally, the main structure of the extracted V-channel image is extracted and the contour edge is preserved by using relative total variation filtering (RTV). The V-channel image is reconstructed to complete image enhancement. Through experimental verification, the regularization coefficient and adaptive parameters are determined. The adaptive image fusion method in HSV space can significantly improve the fusion effect of infrared/visual images in various low brightness conditions, and perform well in detail and edge processing.