Contour Matching in Image-based CAD using B-splines and Digital Geometry

David Durón-Arellano,M. Mehrandezh

Abstract:Recent developments on generating organized Point-Clouds from sequential images taken by cameras is giving the CAD/CAM/CAE practice a quantum leap. One application domain that has gained significant amount of attention is the practice of the fast and precise production of the ergonomic objects for users with the utmost comfort level. This has been paved by the exponential advancement in the computers processing power, and the development of the new algorithms for classifying and merging a huge amount of sequential data in form of organized point-clouds. This is becoming a textbook problem rather than merely a lab curiosity. The main focus of this paper is to explain the automation process on: (1) generating a dense organized point-cloud from sequential images taken by an RGB-D camera, (2) extracting contour lines on the surfaces of interest using the concepts borrowed from digital geometry, and (3) using the information obtained in (2) to construct the objects with seamless ergonomic fit via 3D printing. Some examples are: manufacturing customized gloves, shoes, garments, etc. However, the main focus of this work is on building eye-drop medicine dispenser holders that can be customized for users through a fast, non-intrusive, and semi-automated process. This paper provides a framework on a semi-automated approach for extracting the contours of interest from a digitized CAD model obtained from a point-cloud generated through RGB-D imaging. A 3D model of the subject, i.e., a human face, is first constructed via a mesh using an RGB-D imaging sensor, called structure sensor. The radius of curvature around the scanned eye socket is then calculated along the surface normals on the point-cloud in its vicinity. The points of high deflection, i.e., points corresponding to a sudden change in the radius of curvature, are then used to determine the optimal location of the control points for B-spline fitting. The B-splined contour curve is used as the extrusion base of the eye-drop dispenser holder CAD model. This process can expand to manufacturing of devices such as: eye goggles, wrist bands, shoes, gloves, etc. We plan to incorporate the entire process to a lab module in our CAD/CAM/CAE course which includes all the steps from conception of a design to manufacturing it.