.tf / Ta ~ 2 System Scene Matching Technology

Based on the characteristic of TF/TA2 system, this paper selects two kinds of correlation algorithms based on region intensity as the study object. The two kinds of correlation algorithms are typical SSDA algorithms and Normalized Intensity Combination method (NIC). Then this paper focuses on the capability of restraining the noise, the capability of matching precision and the capability of real time matching. Based on above three guidelines, this paper make a lot of theoretical analysis and example, then a contrast conclusion is given. The contrast conclusion proves the advantage on applying NIC algorithms to TF/TA2 system. At the same time based on the real-time problems of NIC algorithms, this paper introduces elamination search and genetic algorithms into NIC algorithms, and some improvement methods are put forward. Those methods mainly as follows. First, The matching speed is improved through the distinguish difference of two images in delamination search. Second, Niche algorithms is introduced into genetic algorithms in order to solve prematurity problem.