The increase in awareness of the need to improve quality control on part machining efficiency has led to a great deal of research aimed at cutting tool geometry analysis. This paper presents a framework of preprocessing point-based data and extracting parameters after feature detection and data segmentation for cutting tool inspection, assuming unorganized measurement data. The data processing method, including data decimating, smoothing, normal and curvature estimating, denoising, sorting, as well as re-sampling, are exploited to meet the demands for high quality, data simplification for geometric analysis. We will discuss the geometry analysis for parameter extraction, including key feature point detection and key area segmentation based on general reverse engineering solutions and specific cutting tool characteristics. Based on the presented simplification methods using virtual slicing and rotary axial projection data, some cutting tool dimensional parameters can be extracted directly. Alternately, based on 2D points on a given cross section, a plurality of curves can be generated, and optimized by minimizing deviations between the set of points and the plurality of curves. Section parameters can then be extracted from the optimized curves. Furthermore, the methods and processes of multi-section based spatial parameter extraction will be illustrated. This paper presents experimental results and field tests. The experimental results show that the preprocessing is very robust and the parameter extraction results agree with what is expected.

Date Published: 10 September 2009
PDF: 12 pages
Proc. SPIE 7432, Optical Inspection and Metrology for Non-Optics Industries, 74320C (10 September 2009); doi: 10.1117/12.828890

Published in SPIE Proceedings Vol. 7432:
Optical Inspection and Metrology for Non-Optics Industries
Peisen S. Huang; Toru Yoshizawa; Kevin G. Harding, Editor(s)