Measuring of temperature in steel bar using machine vision system and genetic algorithm in variable conditions

The temperature measurement of steel bar is a significant parameter in the hot-rolled steel processing. Furthermore, measuring of steel bar temperature at real time requires an automatic intelligent electronic system such as machine vision system. In this study, a machine vision system is proposed (designed)to provide appropriate images of the steel bar in furnace, process images based on discrimination thresholds and to extract feature (the temperature measurement of steel bar). The threshold limit value for three different temperature ranges; (1200°C -1500°C), (900°C -1200°C) and (600°C -900°C) are assumed to be 1, 2 and 3, respectively, to investigate and analyze variable lighting conditions in HSI color space using genetic algorithm (GA HSI). The performance analysis of proposed GA HSI reveals that color image segmentation by GA HSI and cluster analysis methodhave the same performance. Therefore, this method can overcome the effect of lighting conditions with acceptance of an error range.