In temperature measurement methods of opencast coal stack, the artificial temperature measurement is considered as a commonly used method. However, the artificial temperature measurement has some defects, such as low reproducibility, limited measuring positions and heavy workload. In order to overcome these defects, this paper provided a large-scale, all-weather, multi-target monitoring and effective temperature monitoring method and an internal temperature layered algorithm. Furthermore, an infrared internal temperature detection system was designed and implemented. According to the inner temperature detection method, the coal stack was divided into inner layer and outer layer according to the temperature influence by external environment. The bounding surface was defined as the junction between inner and outer layer of coal stack. The outer layer was varied with environment, and then the inner layer was relatively stable. The surface temperature of coal stack was captured by far-infrared image instrument, and weather conditions were measured by small weather station. The temperature correction algorithm is designed to reduce surface temperature measurement error caused by the atmospheric effect and reflected radiation. Then, the corrected temperature is used as the true surface temperature to calculate internal temperature by layered algorithm. The temperatures of bounding surface is calculated by the heat balance equation of coal stack, and the temperature of inner layer is calculated through exponential relationship between the temperature of bounding surface and the distance from calculated position to the bounding surface. The experimental results of the proposed method indicate that the internal temperature calculation error is 10.36%, and satisfied the temperature detecting practical demand.