摘要
在护理机器人的研究中,摄像机标定是护理机器人伺服控制的前提和重要的步骤。建立包含摄像机内参数、外参数及畸变系数的非线性模型,在优化摄像机标定参数的过程中,设计了一种在PSO算法的基础上融合蒙特卡罗算法非线性优化的MPSO(Modified Particle Swarm Optimization Algorithm),在精度、稳定性和全局搜索能力等方面与一般算法有明显提高。对所求的初始参数进行非线性优化,得到最终的摄像机参数精确值。在Matlab环境下进行仿真,实验结果表明与传统标定算法相比,非线性模型下的MPSO算法标定精度较高。根据摄像机标定结果,可实现护理机器人快速精准定位,并准确地进行视觉伺服控制。
Camera calibration is the premise and important step of the nursing robot servo control. In the process of optimization of camera calibration parameters, this paper established a nonlinear model includeing the camera inter- nal parameters, the external parameters and distortion coefficients, and designed a Modified Particle Swarm Optimiza- tion Algorithm （MPSO） for nonlinear optimization. Based on PSO algorithm and monte carlo algorithm, the accura- cy,＂ stability and global search ability are obviously enhanced compared with the traditional calibration algorithm. Nonlinear optimization was carried out to meet the desires of initial parameters, and the final accurate camera parame- ters were obtained. Based on the Matlab simulation, the experimental result shows that calibration precision of the nonlinear model of the MPSO algorithm is better than traditional calibration algorithm. According to the results of camera calibration, servoing control of nursing robot visual is more accurately, which achieves fast and accurate orien- tation.
出处
《计算机仿真》
CSCD
北大核心
2014年第1期421-424,共4页
Computer Simulation
基金
国家国际科技合作项目(2011DFA10440-3)
关键词
摄像机标定
非线性模型
改进粒子群算法
护理机器人
Camera calibration
Non-linear model
Modified particle swarm optimization algorithm
Nursing ro-bot
