Data Collection and Model Construction Methods for Reverse Engineering
The process of creating a CAD model from an object is mainly made up of two steps: the data collection through digital measurement and the construction of parameterized and revisable model. This paper discusses the measuring process and technical problems of the Coordinate Measuring Machine (CMM) and non-contact sensor. Through comparative analysis, we determine the application scope of those approaches in measuring different dimensions of the same objects considering the time efficiency and tolerance requirement. This paper divide the objects into two categories: freeform feature objects and regular feature objects. As for the freeform feature objects, people could fit wrap-around B-spline surfaces to construct the model. Regular feature objects for mass produce contain machined surfaces which should be precisely measured and modeled. The model of regular feature object should be constructed by three-dimensional modeling software, so that it is parametric and revisable for changing and improving the original design. Sizes and position of important surfaces of the model are acquired from CMM, and those of non-important features are fitted though point cloud processing. Some profile can’t be measured directly from CMM but should be precise, so this paper proposed two methods to construct profile line and analyze error by comparing it with point cloud.
Guozhong Chai, Congda Lu and Donghui Wen
L. Y. Jiang and Z. B. Wang, "Data Collection and Model Construction Methods for Reverse Engineering", Advanced Materials Research, Vols. 102-104, pp. 189-193, 2010