Colour Space Mining For Industrial Monitoring
The effectiveness of colour spaces for the encapsulation of multivariate and distributed source information has led to increasing interest in their deployment in industrial monitoring technology. Other advantageous features are the availability of several hardware technologies that allow early transformation of a variety of monitored quantities into colour information, the compression of data at an early stage, and the familiarity of colour as a means for human assimilation of quantitative information.
To fully realise the potential of colour space representation for enhancing information gathering and enrichment demands its integration with the data mining process. In monitoring applications this must include a capacity for real-time distributed processing and delivery. This contribution describes the relevant characteristics of the representation of information in colour spaces and some of the ways in which it is being combined with data mining methods to develop industrial monitoring solutions. These are illustrated by particular examples from the energy and manufacturing industries of clustering, visualisation and statistical analysis for colour space information.
KeywordsColour Space Virtual Sensor Data Mining Process Colour Science Main Pump
Unable to display preview. Download preview PDF.
- Atlas, L. and Jones, D., “National Science Foundation Workshop on Signal Processing for Manufacturing and Machine Monitoring”, Draft 1.2 (11/15/96) of Final Report to the National Science Foundation, Alexandria, VA, March 14 - 15, 1996 http://isdLee.washington.edu
- Connor, F.R., Signals, Edward Arnold (Publishers) Ltd. (2nd ed.), 1982Google Scholar
- Cosgrave, J.A., Russell, P.C., Hall, B.S.B. and Jones, G.R., “Optoacoustic monitoring of electric arcs in high voltage circuit breakers”, Int. Conf. on Switching Arcs, Xian Jiaotong University, People’s Republic of China, pp. 598–605, 1997Google Scholar
- Fayyad, U.M., “Editorial”, in Data Mining and Knowledge Discovery, pp. 5–10, Kluwer, 1997Google Scholar
- Gabor, D., “Theory of Communication”, J. IEE, 93 (3), 429–457, 1946Google Scholar
- Gasquet, C. and Witomski, P., Fourier analysis and applications, Springer-Verlag Inc., 1999Google Scholar
- Jones, G.R, “Plasma monitoring using chromatically processed optical signals”, Proc. XXI Int. Conf. on Phenomena in Ionised Gases Proc. III (Invited Lectures) (Bochum), pp. 24–33, 1993Google Scholar
- Jones, G.R., “Electric arc monitoring using intelligent optical fibre systems”, Proc. XI Int. Conf. on Gas Discharges and their Applications (Tokyo), vol. 2, pp. 504–512, 1995Google Scholar
- Jones, G.R., Russell, P.C., Vourdas, A., Cosgrave, J., Stergioulas, L.K. and Haber, R., “The Gabor transform basis of chromatic monitoring”, Meas. Sci. Technol., 11, 1–8, 2000Google Scholar
- Levkowitz, H. and Herman, G.T., “GLHS: A generalized lightness, hue, and saturation color model” CVGIP: Graphical Models and Image Processing, 55 (4), 271–285, 1993Google Scholar
- Quinlan, J.R., “Induction of decision trees”, Machine Learning, 1, 81–106, 1986Google Scholar
- Russell, P.C., Craven, A., Deakin, A., Cooke, R., Furlong, S., Spencer, J.W. and Jones, G.R, G.R., “Intelligent chromatic monitoring of industrial plant”, Sensors and their Applications X, Eds. N M White and J T Augousti, Proceedings of the Tenth Conference on Sensors and their Applications, Cardiff, Institute of Physics Publishing, Bristol, pp. 245–250, 5–8 September 1999Google Scholar
- Russell, P.C., Brazier, K.J., Deakin, A., Cooke, R. and Jones, G.R., “Holistic Monitoring of Power Plant”, Intelligent Systems and Applications (ISA 2000), Wollongong, Australia, 11–15 Dec. 2000Google Scholar
- Stergioulas, L.K., “Time-frequency methods in optical signal processing“, PhD Thesis, University of Liverpool, UK, 1997Google Scholar