Electricity transmission equipment in forests poses a great threat to the ecosystem due to the risk of fire, and the rapidity of forest fires or electricity equipment fires makes it difficult to make quick and effective judgments when fires start and to prevent miscalculations in cases such as electrical sparks. This paper designs an intelligent fire protection system that can realize automatic fire extinguishing. This paper focuses on the study of the jet trajectory of the fire extinguishing agent after leaving the fire cannon, and combines the fluid mechanics and particle kinematics to theoretically derive the trajectory and realize the trajectory prediction. Adding air resistance to the trajectory model makes the trajectory prediction more accurate. At the same time, the target detection algorithm in deep learning is used to accurately detect flames, and small flames in the environment can be quickly detected. Using computer vision technology to spatially locate the fire source, obtain the three-dimensional coordinates of the fire source, and calculate the pitch angle and horizontal rotation angle. Compared with existing methods, our proposed method can use deep learning visual detection algorithms to quickly detect flames and combine with other devices to extinguish the fire, constructing a complete fire prevention system, which has stronger significance in practical applications.