On-Line Seam Detection in Rolling Processes Using Snake Projection and Discrete Wavelet Transform

[+] Author and Article Information
Jing Li, Jianjun Shi

Department of Industrial and Operations Engineering, The University of Michigan, 1205 Beal Avenue, Ann Arbor, MI 48109-2117

Tzyy-Shuh Chang

 OG Technologies, Inc., 4300 Varsity Drive, Suite C, Ann Arbor, MI 48108

Here, the “time” is not clock time, but is defined in a broad sense, as the order of the pixels in the sequence.

J. Manuf. Sci. Eng 129(5), 926-933 (May 03, 2007) (8 pages) doi:10.1115/1.2752519 History: Received December 01, 2004; Revised May 03, 2007

This paper describes the development of an on-line quality inspection algorithm for detecting the surface defect “seam” generated in rolling processes. A feature-preserving “snake-projection” method is proposed for dimension reduction by converting the suspicious seam-containing images to one-dimensional sequences. Discrete wavelet transform is then performed on the sequences for feature extraction. Finally, a T2 control chart is established based on the extracted features to distinguish real seams from false positives. The snake-projection method has two parameters that impact the effectiveness of the algorithm. Thus, selection of the parameters is discussed. Implementation of the proposed algorithm shows that it satisfies the speed and accuracy requirements for on-line seam detection.

Copyright © 2007 by American Society of Mechanical Engineers
Your Session has timed out. Please sign back in to continue.



Grahic Jump Location
Figure 1

A portion of a sensing image from a bar-rolling process

Grahic Jump Location
Figure 3

(a) Two ridge-based false positive sub-images (left: containing two dark strips; right: containing one dark strip), (b) a mark-based false positive sub-image

Grahic Jump Location
Figure 9

T2 control chart at k=8, s=5 for the testing dataset

Grahic Jump Location
Figure 8

Chi-square quantile-quantile plot of features (x-axis: chi-square quantiles; y-axis: squared Mahalanobis distances (14) of features; the labels for x- and y-axes are omitted)

Grahic Jump Location
Figure 7

Flow chart of the procedure for selecting snake-projection parameters

Grahic Jump Location
Figure 6

(a) Sequences of the ridge-based false positive sub-images in Fig. 3 by snake projection (k=8,s=5) (top: corresponding to the sub-image with two dark strips; bottom: corresponding to the sub-image with one dark strip), (b) sequence of the mark-based false positive sub-image in Fig. 3 by snake projection (k=8,s=5)

Grahic Jump Location
Figure 2

A seam sub-image

Grahic Jump Location
Figure 5

Sequence of the seam sub-image in Fig. 2 by snake projection (k=8,s=5)

Grahic Jump Location
Figure 4

Procedure of the snake-projection-wavelet algorithm



Some tools below are only available to our subscribers or users with an online account.

Related Content

Customize your page view by dragging and repositioning the boxes below.

Related Journal Articles
Related eBook Content
Topic Collections

Sorry! You do not have access to this content. For assistance or to subscribe, please contact us:

  • TELEPHONE: 1-800-843-2763 (Toll-free in the USA)
  • EMAIL: asmedigitalcollection@asme.org
Sign In