摘要
自适应遗传算法与传统遗传算法不同在于其交叉概率和变异概率随个体的适应度值而变化,能够避免陷入局部最优值。本文针对机器人视觉伺服控制中双目图像匹配问题采用了一种基于自适应遗传算法的同名点匹配方法,根据适应度在算法运行的不同阶段的变化情况来自动修正交叉概率和变异概率,以获得全局最优解。仿真结果表明:该算法能减少匹配算法的计算量,提高了匹配的速度和精度。
The difference between adaptive genetic algorithm(AGA)and conventional genetic algorithm is in that the probabilities of crossover and mutation of AGA vary with the fitness values,and AGA can avoid getting into the local optimum.A correspondence points matching approach based on adaptive genetic algorithm is used in this paper,aiming at the problem of binocular images matching in robot visual servo control.AGA updates the probabilities of crossover and mutation according to the fitness variation in this algorithm′s different operation stages to obtain the global optimal solution.The simulation results show that the approach can reduce the matching calculation amount,improve the matching speed and precision.
出处
《常州大学学报(自然科学版)》
CAS
2010年第2期53-57,共5页
Journal of Changzhou University:Natural Science Edition
基金
江苏省高校自然科学基础研究项目资助(07KJD510038)
关键词
自适应遗传算法
同名点匹配
立体视觉
adaptive genetic algorithm
correspondence points matching
stereoscopic vision