Image Matching Method Based on Hausdorff Distance of Neighborhood Grayscale

摘要

As for the problems between the visual and infrared images which have large differences in gray value caused by different imaging mechanism, inconsistent contour and the low matching probability of traditional matching methods based on gray or feature, gray information of visual/infrared images is introduced after researching a variety of Hausdorfff distance algorithms. Image matching method based on Hausdorfff distance of neighbor grayscaling information is proposed. Based on calculating the similarity of edge feature points, the calculation of neighborhood grayscaling variance is added in this method, effectively solving the low probability problem caused by different edge of infrared image in Hausdorff distance matching algorithms. By the simulation results of visual and infrared images matching, it shows that in various conditions, compared with the conventional Hausdorff distance method, this algorithm has effectively improved the matching effect under different light effects and the anti-jamming noise.