Fast image correlative matching based on genetic algorithm

Shuo Hu,Ming Zhu,HuaJun Song
2005-01-01
Guangxue Jishu/Optical Technique
Abstract:In the view of the problem of conventional correlative matching algorithm, a fast image correlative matching algorithm based on genetic algorithm was presented. According to the property of image datum, new coding was applied, new mutation and crossover operator were given. A novel initialize population approach named sampling approach was applied, and the strategy of competition was introduced, those reduced iteration step and computational cost. The experimental results show the proposed algorithm is effective and efficient. In addition, The computation time is nearly constant for matching each frame of images in the proposed algorithm, which brings convenience to the practical application.
What problem does this paper attempt to address?