Color contrast enhancement for color night vision based on color mapping

The increasing availability and deployment of imaging sensors operating in multiple spectral bands has led to a large research effort in image fusion, bringing a lot of pixel-level color fusion algorithms. Color constancy and color contrast are two important topics for color image fusion. However, there is no work considering these two aspects in the meanwhile. The paper focuses on enhancing the color contrast between targets and their backgrounds and maintaining the color constancy. Firstly, a group of visible and infrared images are rendered with natural colors with color contrast enhancement algorithm. Then a color reference palette is derived from source images and their corresponding fused results. To enrich this palette, the nearest Euclidean distance determines the absent colors. The complete palette can be deployed in real-time to different multi-band image sequences of similar scenes. Experimental results based on different data sets show that the targets are popped out with intense colors while backgrounds present natural color appearance. Some objective analysis also shows that the presented method not only inherits more detail information from source images but also brings bigger color distance when comparing with the existing algorithms.

► Combination of color contrast enhancement and reference palette techniques. ► Missing values are interpolated using the nearest Euclidean distance. ► The method is validated through objective metrics.