Effect of Imaging Distance and Chicken Body Size on Infrared Thermal Camera Accuracy in Body Temperature Measurement

Abstract

The accurate monitoring of body temperature is critical to poultry health and welfare. Rectal thermometry, the conventionally employed method, is invasive and stressful. Infrared thermography (IRT) offers a non-invasive alternative, but its accuracy may be influenced by body size and camera-to-object distance. This study evaluated the efficiency of thermal imaging compared with rectal thermometry in measuring chicken body temperature, with a focus on the effects of body size and measurement distance. A cross-sectional, repeated-measures design was applied to ninety clinically healthy Buff Sussex chickens (n = 30 per size group: small, medium, and large). Each bird was imaged from three distances (50, 75, and 100 cm) by using a thermal camera (FLIR C5®), with rectal temperature (Omron MC-246®) serving as the reference, so that a total of 270 paired observations were analyzed. Agreement was assessed using Bland–Altman bias and limits of agreement (LOAs), root mean square error (RMSE), mean absolute error (MAE), Pearson correlation (r), and Lin’s concordance correlation coefficient (CCC). The results showed that rectal temperatures were consistent with the normal physiological range reported for healthy chickens (40.43–41.95 °C) and that thermal imaging showed systematic underestimation, particularly in small birds (bias of up to −2.65 °C and RMSE of 2.72 °C at 100 cm) and medium-sized birds (bias of −0.73 to −1.39 °C), with weak concordance (CCC ≤ 0.16). Measurements in large birds demonstrated the smallest bias (−0.76 to +0.16 °C), lower errors (MAE of 0.73–0.89 °C), and stronger correlations (r = 0.56–0.71), indicating more reliable agreement. Distance influenced accuracy, with underestimation increasing at 75–100 cm, especially in smaller birds. Therefore, thermal imaging cannot fully replace rectal thermometry for individual-level assessment in chickens due to systematic underestimation, especially in small birds and at greater distances. However, it shows promise as a rapid, non-invasive flock-level screening tool in larger chickens when used at optimal distances (50–75 cm). The integration of thermal imaging into precision livestock farming and future farm models may enhance welfare-friendly, automated health monitoring in poultry systems.