A real time 3D surface measurement system using projected line patterns.

Abstract

This thesis is based on a research project to evaluate a quality control system for car component stamping lines. The quality control system measures the abrasion of the stamping tools by measuring the surface of the products. A 3D vision system is developed for the real time online measurement of the product surface. In this thesis, there are three main research themes. First is to produce an industrial application. All the components of this vision system are selected from industrial products and user application software is developed. A rich human machine interface for interaction with the vision system is developed along with a link between the vision system and a control unit which is established for interaction with a production line. The second research theme is to enhance the robustness of the 3D measurement. As an industrial product, this system will be deployed in different factories. It should be robust against environmental uncertainties. For this purpose, a high signal to noise ratio is required with the light pattern being produced by a laser projector. Additionally, multiple height calculation methods and a spatial Kalman filter are proposed for optimal height estimation. The final research theme is to achieve real time 3D measurement. The vision system is expected to be installed on production lines for online quality inspection. A new 3D measurement method is developed. It combines the spatial binary coded method with phase shift methods with a single image needs to be captured.