LOW-COST CALIBRATION METHOD FOR THE INFRARED CAMERA

Infrared (IR) or thermal cameras are being increasingly used in livestock research and management. An IR camera's accuracy is specified over its entire surface temperature measurement range, whereas in livestock research and management, a narrower range suffices. A camera's accuracy could be higher in a narrower range of temperatures. Hence, a novel low-cost method was used to calculate the FLIR E8 camera's accuracy in a range of 24 degrees C to 37 degrees C, representative of surface temperature of poultry birds. Sensitivity analyses were also performed to evaluate the impact of three user spec-ified parameters, namely, emissivity (epsilon), distance between camera and surface (d), and reflected air temperature (tair). A linear regression model was used to correct the camera's absolute error of 2.8 degrees C (greater than its published error). How-ever, the camera possessed precision and hence, repeatability. The IR camera was moderately sensitive to epsilon, and slightly sensitive to tair and d, but its error could increase with the difference between the measured and assumed tair values. Attention is required to accurately characterize epsilon and tair. This simple calibration method can reduce cost and could improve accuracy in a narrower temperature range than the IR camera's published range, which could be useful for applied research.