Structured Light Techniques for 3D Surface Reconstruction in Robotic Tasks

Abstract

Robotic tasks such as navigation and path planning can be greatly enhanced by a vision system capable of providing depth perception from fast and accurate 3D surface reconstruction. Focused on robotic welding tasks we present a comparative analysis of a novel mathematical formulation for 3D surface reconstruction and discuss image processing requirements for reliable detection of patterns in the image. Models are presented for a parallel and angled configurations of light source and image sensor. It is shown that the parallel arrangement requires 35% fewer arithmetic operations to compute a point cloud in 3D being thus more appropriate for real-time applications. Experiments show that the technique is appropriate to scan a variety of surfaces and, in particular, the intended metallic parts for robotic welding tasks.KeywordsPoint CloudStripe PatternCalibration PlaneRobotic TaskMaximum Span TreeThese keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.