Computer vision system for high temperature measurements of surface properties

Abstract

Recently, computer vision systems have become very popular. They are of great importance in almost every field of science, engineering and industry. Present-day vision systems allow to obtain information that is normally not distinguishable by humans. It is possible due to use of appropriate digital image processing and analysis algorithms. The paper explains importance of proper image processing algorithms selection for computer vision applications. Particular industrial image quantitative analysis systems are considered. Computerized system for high-temperature measurements of surface properties is used as an example. The system is capable of measuring wetting angle and surface tension of metals, alloys and other materials (e.g. glass) in temperatures up to 1,800°C. A brief description of the system is given. Particularly, attention is paid to preprocessing algorithms. They consider not only typical factors that usually accompany digital images founded analysis but also specificity of images obtained during the measurement process as well. Correction of factors arising from CCD camera electronic components and reduction of aura (glow that appears around specimen in high temperatures) affects with high quality image segmentation. In consequence the accuracy of surface parameters determination is increased.