A Complementary Sensor Approach to Reverse Engineering

Abstract

A complementary sensor technique for reverse engineering objects that are represented by a three-dimensional (3D) cloud data set is reported. The research focuses on objects whose surface form is manifest as a set of distinct free-form surface patches, each of which is enclosed by a boundary. The method incorporates three stages: (1) laser scanner-based digitization of all the free-form surface patches, (2) touch probe-based digitization of the surface patch boundaries, and (3) modeling of both data sets to create a complete B-spline curve and surface representation of the object. The patch boundary data, defined by the touch probe, is employed to segment the free-form surface data into the constituent patches. Furthermore, the boundary data is incorporated within a B-spline surface fitting process to constrain the boundaries. The two sensors functionally complement each other; the range sensor provides the required dense resolution of 3D points on the free-form surfaces whereas the touch probe accurately defines the patch boundaries. The method is ideal for objects comprised of both functional engineering features, e.g. bearing holes or precise mounting locators, and aesthetic features, such as hand grips or part covers. The touch probe is also ideal for digitizing boundaries where occlusion prevents the use of an optical digitizer. The laser-based sensor has an accuracy specification of 50 microns (over a 40-mm depth of field) whereas the touch probe is accurate to 4 microns over a 25-mm measurement length. An example part is modeled that has multiple free-form patches (defining the part’s outer cover) that require a large cloud data set for complete coverage. The corresponding patch boundaries accurately define the location of critical part mounting locations that require the touch probe’s precision.