A Focus-Measurement Based 3D Surface Reconstruction System for Dimensional Metrology

Abstract

Abstract Image focus is an important parameter which depends on various other parameters of an imaging system. One among them is the distance of the object plane from the lens. This parameter is referred as the depth in computer vision literatures. Since depth affects the focus level of pixels in an image, focus measurement can be used to interpret the depth of points of an object. The paper presents the details of the novel indigenously developed automated measurement system which can perform high-precision movements based on feedback control. The measurement system is inspired by classical systems used in techniques such as shape from focus. This system fundamentally being a mechatronics system integrates a high-precision multi-stage translation mechanism which is actuated with a servo motor with sensors such as camera, linear encoder, optical limit switches. The hardware interface is developed for automated measurement through a USB data acquisition system. The paper presents the methodology of surface reconstruction and detailed description of the various challenges encountered in the integration. The surface reconstruction is performed for various millimetres and sub millimetres level engineered specimens. The paper presents the future direction of research and also various potential application possibilities for such measurement setup.