摘要
目前智能清洁机器人的清洁覆盖率的测试主要采用单目视觉方法,针对其测量范围小、获取信息不完整、测量精度低等缺点,设计并实现了基于双目视觉特征跟踪算法的清洁机器人清洁性能测试系统。该系统采用摄像机、智能清洁机器人为硬件平台,Ground Truth System为软件平台,运用SURF算法提取具有高鲁棒性的特征点,在后续帧中运用KLT匹配算法对特征点进行稳定跟踪,结合机器人运动路径和机器人参数实现了智能清洁机器人清洁覆盖率的测量。实验证明,该方法对智能清洁机器人的清洁覆盖率测量是快速有效的。
The cleaning coverage performance test system of intelligent cleaning robot mainly based on monocular vision. In order to overcome its shortcoming as small measuring range,incomplete information and low measurement precision,a design and implementation of cleaning coverage performance test system of intelligent cleaning robot based on binocular vision feature tracking algorithm is introduced. The system uses the camera and intelligent cleaning robot as hardware platform,Ground Truth System is used as software platform. Feature point of high robustness extraction using the SURF algorithm,using the KLT matching algorithm for feature point tracking in subsequent frames,measuring cleaning coverage performance of intelligent cleaning robot through the path and parameter of robot. The experiment indicates that the method of intelligent cleaning robot cleaning coverage measurement is fast and effective.
出处
《激光杂志》
北大核心
2015年第6期130-134,共5页
Laser Journal
基金
重庆市教委科学技术研究项目(编号:KJ120519)
重庆市教委科学技术研究项目(编号:KJ130512)
