摘要
在结构光高精度测量与生产线自动化结合的过程中往往要求实时性。为了有效提取测量物体的点云信息以及提高运算效率,利用面阵相机与投影仪像素点有序的特性,提出一种针对有序点云的快速去飞点算法。首先,根据点云数据估计投影矩阵,通过估计的投影矩阵把点云投影到一个像平面,然后基于滑动窗口把像平面上的每一个点最近邻搜索k个临近点,最后根据概率阈值并行地统计分析某点的k邻近点,符合条件的作为飞点去除。该算法易于工程实现,简单有效。实验结果表明,该算法具有良好的点云滤波效果和计算效率。在一定条件下可以达到工业化测量的精度和速度要求。
There are real-time requirements in the process of applying the structured light high-precision measurement to the production line automation.In order to effectively extract the point cloud information of the measured object and improve the computational efficiency,a fast outlier removal algorithm for the organized point cloud is proposed that using the feature of pixel order of array camera and projector.Firstly,the point cloud is projected to an image panel according to estimated projection matrix.Then,the k-nearest neighbor algorithm is used to search k neighboring points of each point in the image panel based on sliding windows.Finally,parallel removing a point when its k neighboring points are not satisfy some statistical properties to the probability threshold.The algorithm is easy to implement in engineering,simple and effective.The experimental results show that the algorithm has good point cloud filtering effect and computational efficiency,and accuracy and speed requirements of industrial measurement can be met under certain conditions.
作者
罗国强
左文涛
陈家益
LUO Guo-qiang;ZUO Wen-tao;CHEN Jia-yi(College of Information Engineering,Guangzhou Vocational and Technical University of Science and Technology,Guangzhou 510550,China;College of Engineering,Guangzhou College of Technology and Business,Guangzhou,510850,China;College of Biomedical Engineering,Guangdong Medical University,Zhanjiang,524023,China)
出处
《控制工程》
CSCD
北大核心
2022年第1期129-136,共8页
Control Engineering of China
基金
2021年广东省普通高校特色创新类项目(2021KTSCX179)
广州科技职业技术大学2021年校级课题项目(2021ZR05)
广东省教育厅重点科研平台项目(2017GWTSCX064)
广东省医学科研基金项目(B2018190)
广东医科大学科研基金项目(GDMUM201815)。
关键词
结构光测量
点云滤波
去飞点
有序点云
并行计算
Structured light measurement
point cloud filtering
outiler removal
organized point cloud
parallel computing
