摘要
提出了一种用于图像匹配的混沌搜索算法,首先将目标搜索、匹配、识别的过程转化为函数优化问题,然后运用混沌优化算法对该问题进行寻优,从而获得最佳匹配点。仿真结果表明,在保证寻优性能的情况下,可以减少匹配次数,提高识别速度,满足电视跟踪系统实时性的要求。
A chaotic search algorithm is proposed for image matching. The process of target search, recognition and matching is transformed to the function optimization problem, which is resolved by a chaotic optimization algorithm in order to acquire optimal match point. The simulation results prove that number of times of matching may be reduced and identification may be accelerated, and requirements to TV tracking system are satisfied under the circumstances of ensuring optimizing performances.
出处
《电视技术》
北大核心
2005年第2期70-72,共3页
Video Engineering
基金
国家自然科学基金(10402003)
北京理工大学基础研究基金(BIT-UBF-200301F10)
国家部委预研项目(404050201)
关键词
图像匹配
混沌优化
电视跟踪
image matching
chaotic optimization
TV tracking system
