基于点云骨架的带电作业机器人三维重建系统设计OA
DESIGN OF 3D RECONSTRUCTION SYSTEM FOR LIVE-LINE WORKING TELEOPERATION ROBOT BASED ON POINT CLOUD SKELETONIZATION
为提高带电作业机器人作业的精准性与稳定性,设计面向带电作业遥操作机器人的三维重建系统.提出一种优化的ROSA(Rotational Symmetry Axis)方法提取高压线模型的一维骨架,以及投影映射方法估计高压线截面半径;基于加权法线,提出最小二乘抛物线参数拟合方法拟合高压线的空间参数方程,以修正和优化骨架.实验结果表明,优化后ROSA算法在噪声和点云缺失下实时性、准确度优于其他传统提取算法,且曲线拟合方法适用不同曲率的曲线,符合机器人作业的要求.
To improve the precision and stability of live-line working robot,a 3D reconstruction system is designed for live-line working teleoperation robot.A skeletonization method based on ROSA method optimization was proposed to shrink line model into one-dimension skeleton shape,and a projection mapping method was used to estimate the sectional radius of high-voltage line.Based on weighted normal,a least squares parabola parameter fitting for high-voltage line space parameter equation was advanced to correct and optimize the extracted 1D skeleton.Experiment results show that the skeletonization method based on ROSA method optimization has higher accuracy and real-time and more robustness than tradition methods,and curve fitting method can apply to curves with diverse curvature and the reconstruction system meets the requirements of robot.
闻知健;李会军;宋爱国
东南大学仪器科学与工程学院 江苏 南京 210096东南大学仪器科学与工程学院 江苏 南京 210096东南大学仪器科学与工程学院 江苏 南京 210096
信息技术与安全科学
高压线三维重建带电作业机器人点云骨架三维点云遥操作目标识别空间抛物线拟合
High-voltage line model reconstructionLive-line working robotPoint cloud skeletonization3D point cloudTeleoperationTarget recognitionParabolic curve fitting
《计算机应用与软件》 2026 (3)
10-16,7
国家自然科学基金项目(62173088)国家重点研发计划项目(2022YFC2009700,2022YFC2009704).
评论