基于辐射测温原理，介绍了红外测温理论和自适应维纳滤波处理方法，为了提高钢水的测温精度，对钢水的红外热图像噪声进行处理。搭建了实验平台，获得了不同温度下的钢水的红外图像，利用自适应维纳滤波对钢水红外图像进行去噪处理。利用Matlab 软件进行仿真，通过对处理前后的红外图像的均方差和峰值信噪比进行比较，发现钢水红外图像的噪声得到了抑制。

Based on the principle of radiation temperature measurement, this paper introduces the theory of infrared temperature measurement and adaptive wiener filtering processing method in order to improve the molten steel temperature measuring accuracy and reduce the infrared thermal image noise of molten steel processing. Having set up the experimental platform, we obtain the different temperature of molten steel in the infrared image, using adaptive wiener filtering for molten steel infrared image denoising processing. Before and after using the Matlab software simulation, and through the processing of infrared images of the mean square error and peak signal to noise ratio are compared, we find that steel infrared image noise has been suppressed and the method for the application of infrared temperature measurement in the molten steel temperature measuring has laid a solid foundation.