Aviation blade inspection based on optical measurement

Abstract

Inspecting the blade by optical method is a meaningful work in manufacturing industry. One common problem encountered is that the scanned point cloud is large-scale and noisy. In this paper, we present a systematic introduction of simplification, smoothing and feature extraction. The moving least square surface is applied to create a geometric deviation, which is used to identify sparse points or excessive deviation points, in order to subdivide and cluster the point cloud. Then, the information entropy in k-neighbourhood is defined to distinguish density difference of blade point cloud. The objective is to smooth point-sampling surface meanwhile preserving high curvature feature. Furthermore, the computation method of single/multi section parameters is presented. Finally, two cases are carried out to demonstrate the feasibility and effectiveness. © 2013 Springer-Verlag Berlin Heidelberg.