A feedback design to a CAD-guided area sensor planning system for automated 3D shape inspection

Abstract

A feedback-based area sensor planning system is introduced in this paper. This system is developed for automated 3D shape inspection. A CAD-guided area sensor planning system had been developed previously. However, physical properties of a part surface are usually difficult be modeled for an area sensor. Shadows, reflections, and image quality are three problems related to surface measurement quality, which cannot be predicted from a CAD model. Feedback from a measured 3D point cloud can be used for dealing with those problems by adding more viewpoints in iterations. In feedback mode, the inspection system generates viewpoints according to point clouds measured in each step. The loop stops when the obtained point clouds satisfy the measurement quality requirements. In this paper, a general framework of such area sensor planning system is introduced. The developed feedback system has been implemented for a robot-aided dimensional inspection system.