Heat stress of broilers in commercial broiler-houses decreases their productivity and thus farm profitability. Climate control systems use sensors measuring the temperature around the broilers, which can be different from the actual body temperature of the broilers. In this research, a method estimating the body temperature of an individual broiler was designed and validated for commercial broiler-houses. The method is based on a low-cost IR camera calibrated in real time. The algorithm for the IR image processing uses lasso regression prediction model. A prototype using this method was built and tested in a research broiler-house over 21 days (age 14–35 days) for 15 broilers. The predicted body temperature was compared with the actual body temperature measured by temperature loggers implanted in the abdominal cavity. The accuracy was ±0.27 °C measuring one broiler every 16 min on average. In an experiment in a commercial broiler house, a discrepancy between the climate control activation and broiler estimated body temperature was shown.

中文翻译：

---

热像仪自动测量肉鸡温度

商业肉鸡舍中肉鸡的热应激降低了它们的生产力，从而降低了农场的盈利能力。气候控制系统使用传感器测量肉鸡周围的温度，该温度可能与肉鸡的实际体温不同。在这项研究中，为商业肉鸡舍设计并验证了一种估算个体肉鸡体温的方法。该方法基于实时校准的低成本红外摄像机。红外图像处理算法使用套索回归预测模型。使用这种方法构建了一个原型，并在一个研究型肉鸡舍中对 15 只肉鸡进行了超过 21 天（14-35 天）的测试。将预测的体温与植入腹腔的温度记录仪测量的实际体温进行比较。精度为±0。27 °C 平均每 16 分钟测量一只肉鸡。在商业肉鸡舍的一项实验中，显示了气候控制激活和肉鸡估计体温之间的差异。