Utilizing Diverse Feature Data for Reconstruction of Scanned Object as a Basis for Inspection

Abstract

Inspection of machined objects is one of the most important quality control tasks in the manufacturing industry. Ideally, inspection processes should be able to work directly on scan point data. Scan data, however, are typically very large scale (i.e., many points), unorganized, noisy, and incomplete. Therefore, direct processing of scanned points is problematic. Many of these problems may be reduced if reconstruction methods exploit diverse scan data, that is, information about the properties of the scanned object. This paper describes this concept and proposes new methods for extraction and processing of diverse scan data: (1) extraction (detection of a scanned object’s sharp features by the sharp feature detection method) and (2) processing (scan data reduction by the geometric bilateral filter method). The proposed methods are applied directly on the scanned points and are completely automatic, fast, and straightforward to implement. Finally, this paper demonstrates the integration of the proposed methods into the computational inspection process.