Research On Fusion Algorithms And Simulation Of Infrared And Visible Images For Space Targets Identification

Multi-source image fusion is implemented by integrating multiple source images with redundant and complementary information into one result with better intelligibility and definition. The fused image can reflect multiple properties of source images. In this way, redundancy will be reduced, while complementary information will be utilized more effectively. Subsequent image processing such as image segment, target detection, tracking and recognition are supported by more information with image fusion.In this dissertation the representative infrared and visible images are selected as the research object and the main research work is focus on the application of infrared and visible image fusion technology in space target identification, including analysis of fusibility, fusion algorithm and its application and performance.Characteristics of imaging of infrared and visible light are studied, and approach for judge of fusibility of infrared and visible images is proposed. Models to judge the validity of imaging sensors are established. The validity of image sensors is judged. The requirements for fusion of infrared and visible images are presented. The models and method are tested and verified by experiments.A sparse representation-based fusion algorithm for infrared and visible images of the same scene is presented. In the method, the source image is represented with sparse coefficients using an over-complete dictionary. Then the coefficients are combined according to the detail information and activity level in the sliding window. Finally, the fused image is reconstructed from the combined sparse coefficients and the over-complete dictionary. In addition, the experiments to verify the algorithm and to study influence of the over-complete dictionary on the proposed algorithm are done, and in-depth analysis of the result is given.The application of image fusion in space targets recognition is studied. When the space targets run into clutter of the earth and the moon, the imaging sensor is partly invalid and the targets in the image may be malformed. Fusion of infrared and visible images is used to solve this problem. The experiments show that the system is more stable and the recognition rate is promoted.