Image Matching Using Genetic Algorithm on GPU
Yongzhen Ke, Yuhao Li, Dandan Li · 2011
Genetic algorithm is widely applied to the field of image matching. But it is very time-consuming to solve some large computational problems with large search space or complex fitness function. Image matching using genetic algorithm has not achieved real-time performance. An improved genetic algorithm based on CUDA with full use of GPU's parallelism was presented in this paper. The experimental results show that improved genetic algorithm on GPU has an excellent performance on acceleration of image marching. Now the GPU supporting CUDA is very popular in personal computer. So the proposed algorithm can be easy to apply other field.