An Medical Image Registration Approach Using Improved Hausdorff Distance Combined with Particle Swarm Optimization

摘要

A new method combined the least trimmed square Hausdorff distance (LTS-HD) with Particle swarm optimization (PSO) is provided. The feature points of the two images are extracted by Harris corner detector to reduce the amount of computation. The affine transform is made between the source image and the target image, and the improved Hausdorff distance is taken as the registration similarity measure. Finally, the translation parameters are calculated by using PSO algorithm. Comparisons are made between the LTS-HD pattern and the MI approach, or the PSO and the Powell optimization on several performance criteria. Experiments results show that the proposed algorithm is effectively and accuracy, and it could reduce the amount of computation largely.