Forest disaster prevention method and system

The invention relates to the technical field of forest management and protection, in particular to a forest disaster prevention method and system, and the method comprises the following steps: carrying out the real-time collection and analysis of heat source, temperature and humidity data in a forest through a deep learning algorithm and a remote sensing technology, and recognizing the abnormal temperature distribution in the forest through a convolutional neural network, and generating a real-time forest fire danger index report. According to the invention, through integration of deep learning and remote sensing technologies, various key parameters in a forest can be captured in real time, the accuracy of early warning of a fire disaster is greatly improved, and the soil moisture and temperature are monitored in real time by using an intelligent sensor network and a support vector machine method, so that the assessment of fire danger is more comprehensive, and the accuracy of early warning of the fire disast