Abstract: A series of forest fire experiments were conducted to systematically evaluate the capability of thermal infrared imaging technology to accurately identify fire points under fire state change, time period difference, and flight altitudes. The dynamic characteristics of the fire point situation were analyzed based on the intensity of thermal infrared radiation and the height of the flame. The findings indicate that thermal infrared technology possesses the capability to accurately identify the fire point. The temporal evolution of thermal radiation intensity exhibits substantial stage characteristics, revealing a quintuple-stage evolutionary pattern in the combustion process of fire sources. The root mean square errors (RMSE) of the flame height formula are 0.0321 m and 0.0402 m. The symmetric mean absolute percentage errors (SMAPE) are 7.07% and 11.29%. And the R² of the linear regression model are 0.9955 and 0.9959. The formula demonstrates notable applicability and accuracy. This study lends further credence to the efficacy and dependability of thermal infrared imaging technology in forest fire monitoring. The technology facilitates dynamic monitoring, early warning systems, and emergency management of forest fires.