The 3D scanning system for the machine vision based positioning of workpieces on the CNC machine tools
The 3D scanning system for the machine vision based positioning of workpieces on the CNC machine tools
复制标题
DOI：
10.1109/mmar.2012.6347906
复制
发表时间：
2012-11
期刊：
影响因子：
--
通讯作者：
K. Okarma;M. Grudziński
中科院分区：
文献类型：
--
作者：
K. Okarma;M. Grudziński
本文讨论了数控机床上工件定位实验系统中用于 3D 扫描系统校准和图像分析的一些算法。扫描的想法基于使用条纹图案方法的摄影测量算法的应用。建立了一个由三个摄像机和三个结构光投影仪组成的实验系统，以便获取代表具有投影光图案的扫描物体的图像。然后分析这些图像以获得代表工件或背景的每个点的深度信息。然而，考虑到相机和投影仪的光学器件引入的失真，良好的精度需要对系统进行适当的校准。在校准、采集图像系列并对其进行分析之后，获得以点云表示的物体的 3D 模型作为结果，该结果应该被过滤并可以拟合到已知模型中。实验中获得的结果也与具有类似相机的可用商业系统的应用效果进行了比较。
In the paper some algorithms applied for the calibration of the 3D scanning system and image analysis in the experimental system for positioning the workpieces on the CNC machines are discussed. The idea of the scanning is based on the application of photogrammetric algorithms using the fringe patterns approach. An experimental system consisting of three cameras and three structural light projectors has been built in order to acquire the images representing the scanned object with projected light patterns. These images are then analyzed in order to obtain the depth information for each point representing the workpiece or the background. Nevertheless, a good accuracy requires a proper calibration of the system considering the distortions introduced by the optics of the cameras and projectors. After the calibration, the acquisition of image series and their analysis, the 3D model of the object, represented as a point cloud, is obtained as the result, which should be filtered and can be fitted into the known model. The results obtained in the conducted experiments have also been compared to the effects of the application of an available commercial system with similar cameras.