Measurement of Critical Dimensions of Cold Roll Formed Sections Using Digital Image Processing

Abstract:

A research investigation is presented which discusses the practicality of using several image processing and knowledge based techniques for the measurement and classification of cold rolled steel sections. Image analysis techniques can be applied to many different applications and assessing the quality and the accuracy of cold roll formed steel sections is no exception. The operations detailed within this paper are both traditional image processing methods and novel neural network based techniques which are combined together to give a bespoke alternative to the manual processing currently employed to test these sections. The results show the suitability of using image analysis and image processing to aid in the quality control of cold steel roll forming and initial tests have demonstrated great potential for this work.

• Periodical: Key Engineering Materials (Volume 473)
• Pages: 949-956
• DOI: https://doi.org/10.4028/www.scientific.net/KEM.473.949
• Online since: March 2011
• Authors: Ian Williams, Bez Shirvani, Jean Michael Mourier
• Keywords: Angle Measurement, Chain-Code, Cold Roll Forming, Digital Image Processing, Edge Detection, Performance Measure, Sheet Metal Forming, Statistical Analysis, Strain Hardening

References

[1] M. Šonka, V. Hlaváč, and R. Boyle. Image Processing Analysis and Machine Vision. Chapman and Hall Publishing, (1986).

[2] J. Canny. A computational approach to edge detection. In IEEE Transactions on Pattern Analysis and Machine Intelligence, volume 8, pages 769-698, (1986).

DOI: 10.1109/tpami.1986.4767851

[3] I Williams. Edge detection of textured images using multiple scales and statistics. PhD thesis, Department of Engineering and Technology, Manchester Metropolitan University, (2008).

[4] Nicholas Bowring, Ian Williams, Christopher Johnson, and Jay Jaiswal. Fatigue crack, squat and wheel burn detection by a multi-scale statistical image processing technique. In British Institute of Non Destructive Testing, (2008).

[5] I. Williams, N. J. Bowring, E. Guest, P. Twigg, Y. Fan, and D. Gadsby. A combined statistical/neural network multiscale edge detector. In 5th IASTED Conference on Visualisation Imaging and Image Processing., pages 91-97, 2005. ISBN: 0-88986-528-0, ref: 480-266.

[6] Adnan A. Y. Mustafa. Boundary signature matching for object recognition. In VI 2001 Vision Interface Annual Conference, pages 72-79, (2001).

[7] H Freeman. On the encoding of arbitrary geometric configurations . In IRE Transactions on Electronic Computers, pages 260 - 268, (1961).

DOI: 10.1109/tec.1961.5219197

[8] I. Williams, D. Svoboda, and N. Bowring. A novel performance metric for grey-scale edge detection. In Proceedings of the International Conference on Computer Vision Theory and Applications., pages 91-97, (2010).

DOI: 10.5220/0002821700910097

[9] M. E. Petersen and E. Pelikan. Detection of bone tumours in radiographs using neural networks. In Pattern Analysis and Application, volume 2, pages 172-183, (1999).

DOI: 10.1007/s100440050026

[10] S. M Bhandarkar, J Koh, and M Suk. Multiscale image segmentation using a hierarchical self-organizing map. In Neurocomputing, volume 14, pages 241- 272, (1997).

DOI: 10.1016/s0925-2312(96)00048-3

[11] L. Piccoli, A. Dahmer, J. Scharcanski, and P O A. Navaux. Fetal echocardiographic image segmentation using neural networks. In Seventh International Conference on Image Processing and Its Applications, volume 2, pages 507 - 511, (1999).

DOI: 10.1049/cp:19990374

[12] D. Skapura. Building Neural Networks. Addison-Wesley, (1996).