基于面结构光的机械工件三维扫描系统设计

Design of 3D scanning system for mechanical workpiece based on surface structured light

摘要:

针对工业机器人对自动化装配过程生产效率的提高以及工件拾取对三维扫描技术的应用需求,设计了能够准确提取机械工件表面点云的视觉系统。扫描系统主要由计算机、投影仪和工业相机构成。基于光学测量和机器视觉的原理,主要研究设计了扫描系统工业相机和投影仪的标定策略、结构光栅编码解码的检测策略以及点云重构的几何策略。对于机械工件三维扫描重构的多余背景平面点云,研究设计了通过随机选取点云并反复迭代构造背景平面实现分割的有效方法。实验结果表明采用面结构光技术,由投影仪投影不同频率的结构光栅在机械工件上,工业相机同步采集被机械工件调制的结构光栅图像,对图像中的光栅条纹进行提取并计算,并利用三角检测算法提取机械工件表面点云的方案具有高准确性,能够有效重构机械工件表面点云。

Abstract:

According to the improvement of production efficiency in industrial manufacturing, especially in the automatic assembly process by the robot and the application requirements of picking up the workpiece to 3D scanning technology, this paper designed a visual system that can accurately extract points cloud of the mechanical workpiece surface. The designed scanning system is mainly made up of computer, projector and industrial camera. Based on the principle of optical measurement and machine vision, this paper researched and designed the calibration strategy of industrial camera and projector, the detection strategy of structured light-coding and decoding, and the geometric strategy of points cloud reconstruction in scanning system. For the redundant background plane points cloud of 3D scanning reconstruction of mechanical workpiece, this paper designed an effective segmentation method by randomly selecting the point cloud and constructing the background plane iteratively. Experimental results show that the scheme of using structured light technology, projecting structured grating at different frequencies on the workpiece by projector, acquiring the modulated raster images of mechanical workpiece synchronously by industrial camera, extracting and calculating the grating in the image, and extracting the points cloud on mechanical workpiece surface by triangular detection algorithm has high accuracy and reconstructs the surface cloud of mechanical workpiece effectively.