Study On Color Fusion Algorithm Of Low-level-light And Infrared Images

Image fusion technology was generated and widely applied in militarysurveillance and other fields with the development of imaging sensors. Color fusion isone of the key researches in this technology for its rich information and taking fulladvantage of the human visual characteristics. The main work of this paper covers aseries algorithm on color fusion of low-level-light and infrared images, and analyzingthem thorough simulation experiments.This pape r first discussed the methodology based on color space mapping, theyobtained the false color fused image by the differences of the input images. In order togive the fused image with natural and realistic appearance, this paper has studied thecolor transfer technique and its fast algorithm, along with the color transfer methodbased on histogram matching. The modified algorithm to enhance the contrast rationbetween background and target was researched, so the fused image could have naturalappearance and pop out targets clearly when fusing according to this procedure. Inaddition, this paper researched the Retine x theory and other color image enhancementmethods, which were intended for the problem of color distortion. The y can deal withthe images acquired at night or under fog conditions. These algorithms can alsoprocess the fuzzy images or images with high-dynamic range, which could be used inthe preprocessing of image fus ion. At last, the performance and applicable occasionsof different color mapping methods, along with the color transfer fusing methods areanalyzed and compared through simulation experiments.