The image fusion based on gray-scale visible light and infrared images lacks color information. Meanwhile, the independency of images’ high-order information is insufficient in transform domain statistics. Therefore, an image fusion approach based on independent component analysis and IHS transform space is presented. The IHS transform can separate the image intensity component and color information effectively. Firstly, the color transfer model of gray-scale image for gray-scale visible light images is erected by IHS transform. Then, according to each component’s independency, the images are fused based on ICA and IHS transform. The final color fused image is got, which fits the need of human vision better. Experimental results validate the effectiveness of the proposed method.