【摘要】 针对棒材弯曲变形缺陷检测问题，提出了一种基于三维点云的棒材弯曲变形检测方法。利用三维传感器多视角扫描，经过降噪、简化、配准操作获取棒材点云数据，RANSAC算法求取棒材轴线方程，依据轴线方程将棒材点云数据坐标变换，基于点坐标信息分割点云，获得多条外形曲线数据，得到棒材弯曲变形数据。棒材检测实验表明：文中方法测量与人工测量最大误差为1.16 mm,对棒材合格判定结果与人工一致，证明了文中方法的有效性。更多还原

【Abstract】 To address the issue of detecting bending deformation defects in bars, a method for measuring bending deformation of bar based on three-dimensional point cloud was proposed.By utilizing a three-dimensional sensor with multiple perspectives, the bar’s point cloud data were obtained through denoising, simplification, and registration operations.The RANSAC algorithm was employed to determine the bar’s axis equation, and based on this equation, a coordinate transformation was applied to the bar’s point cloud data.By utilizing the point coordinate information, the point cloud was segmented to obtain multiple contour curve data, thereby acquiring the bending deformation data of the bar.Through bar detection experiments, the maximum error between the measurements obtained using this method and manual measurements is determined to be 1.16 mm, the result of qualification judgment of bar is consistent with that of manual, demonstrating the effectiveness of the proposed approach.更多还原