CONTRAST BASED IMAGE FUSION

摘要

A system for two color image fusion blending co-registered low light level images in the visible region of the electromagnetic spectrum with thermal infrared images maximizes the information content of the scene by detecting in which of the two image types, IR and visible, there is more structural information and increasing the weight of the pixels in the image type having the most structural information. Additionally, situational awareness is increased by categorizing image information as “scene” or “target” and colorizing the target images to highlight target features when raw IR values are above a predetermined threshold. The system utilizes Red, Green and Blue (RGB) planes to convey different information such that for targets the Red plane is used to colorize regions when raw IR exceeds the predetermined threshold. For scene images, the Green plane provides improved situational awareness due to the above weighted blend of the two image types.