Preliminary Simulation Investigation On Data Fusion Application To Active-Passive Imaging Compound

Because the Ladar and the forward-looking infrared (FLIR) influence the target's reflection of the 10 micron and radiation of 8-12 micron separately, their output information has complementarity and compatibility. Making use of the active-passive compound can improve the ability of target classification and resist interfere.In the background of target classfication fusion, this paper has concentrated on the the classification fusion method D-S evidence theory and the application to the images fusion classification combined with Ladar and FLIR. By doing so, it can provide basis theory and practical algorithm to improve the rate of automatic target classfication.The center of this paper is the active-passive compound image data fusion. It has mainly researched on the target fusion classfication based on D-S evidence theory and the active-passive compound target detection using feature-based fusion algorithm.The emphasis is the D-S evidence theory of the decision-level fusion algorithm and the application to the active-passive compound data fusion, at the base of that, the paper has researched the decision-level fusion algorithm which is based on the D-S evidence theory. Reducing the noise in Ladar range image, finding the target orientation and the choice of the template influence the target fusion classification results. This paper has developed some work corresponding to these aspects, and giving a feasible plan for the active-passive compound target classfication.After that, the paper has introduced a feature-based fusion algorithm which is used in active-passive compound target detection. Detecting the target exactly is the foundation and precondition of the target classification. Through getting four characters such as maximum grey level, contrast, average gradient strength and local variation which are fused by weight can make target detected. When being in various disturbing situation, according to the imaging characters of Ladar and FLIR, single sensor can not detect exactly, but making use of the complement and compatibility of their information can improve target detection accuracy in various situation, which is also to be the foundation for afterward work.