针对传统接触法测温精度低、实时性差等问题，提出了一种基于热辐射理论的红外热像测温模型。首先，根据红外热像仪测温原理，搭建了熔融金属红外测温实验平台，获取了不同温度下熔融金属的红外图像。其次，为了提高测温精度，采用稀疏分解方法去除熔融金属红外图像的噪声，采用最小二乘法和改进遗传算法对提取的图像灰度均值进行灰度-温度的曲线拟合，建立红外图像的灰度值与被测物体表面温度之间的对应关系。最后，利用等温线显示炉内温度整体情况。实验结果表明，此模型可以有效地测量熔融金属的温度，并且测温精度较高，验证了模型的有效性和可行性。

Aiming at the problems of low accuracy and poor real-time performance of the traditional contact temperature measurement methods, this paper proposes an infrared thermal imaging temperature measurement model based on the thermal radiation theory. First, according to the principle of an infrared thermal imaging system, an infrared temperature measuring experiment platform is built, and the infrared images of molten metals at different temperatures are obtained. Then, to improve the precision of temperature measurement, the sparse decomposition method is used to remove the noise in the infrared image of molten metal. The least square method and an improved genetic algorithm are used fit the curve of the gray-level-temperature, which can establish the corresponding relationship between the gray value of the infrared image and the surface temperature of the measured object. Finally, the overall state of the temperature in the furnace is shown using an isotherm. The experimental results show that this model can effectively measure the temperature of molten steel and obtain high temperature measurement accuracy, which verify the validity and feasibility of the model.