Temperature Estimation of Visible Heat Sources by Digital Photography and Image Processing

This paper proposes a novel noncontact temperature measurement technique using a consumer-grade digital still camera. The images of various visible heat sources are captured. Using color image segmentation, source zones are identified. The actual temperature is measured by placing a miniature thermocouple in each zone. The color temperature correlation is established by applying various analytical techniques. Different equations are formulated for different zones observing color dominance in the respective zone. The results indicate that the image-based measurements agree with the corresponding contact measurements within an error range of about +/-5% for most of the sources. The technique is successfully used for measuring the temperature distribution of different industrial applications, like muffle furnace, salt bath furnace, induction furnace, etc.