Research On Key Technique Of Terrain Matching Based On Image

Terrain contour matching guidance system is an important part of theprecision-guided weapons, which is used to control the aircraft navigation and seektarget. Terrain matching technology based on image has been widely used in aircraftguidance systems, with independent, high accuracy. But the image matching iscompute-intensive operations, and it is difficult to achieve real-time, reliability,embedded. So the countries in the world equipped with such weapons, while alsocontinue to improve terrain matching algorithm to improve guidance precision. How toimprove image-based terrain matching algorithm efficiency has become a hot issue.There are two ways to improve the matching efficiency, one is to minimize the numberof search locations, and the other one is to simplify the similarity measurementalgorithm.Through the research on the theory of terrain matching and particle swarm, thispaper proposes an improved particle swarm algorithm, and it can speed up theoptimization speed, solve the problem of prematurity. Then similarity measure terrainmatching algorithm is improved to reduces matching calculation. Finally, an improvedterrain matching algorithm based on image is proposed. The algorithm fitness functionand initial population problem has been discussed. The algorithm can improve theterrain matching problem In the optimization path and similarity measure algorithm.Simulation results show the improved terrain matching algorithm can solve theterrain matching problem, with a small amount of computation, fast speed, highprecision, and not easy to fall into prematurity. It can increase the matching accuracyand efficiency, so it is a kinds of effective terrain matching algorithm.However, in this paper, the parameters of PSO reference from references, orthrough a large number of experiments. There is no strict mathematical theory as a basis.In the follow-up study research, the algorithm parameter settings requires further study.The similarity measurement algorithm has a lot of research directions, which are worthyof study.