摘要
森林火灾是一种严重的自然灾害,且我国地域辽阔、森林面积广阔,对于森林火灾的预警监测需求量巨大。针对我国森林火灾问题,从扩大森林火灾监测范围、提升森林火灾识别响应速度的角度出发,设计了一种基于无人机图像拍摄、利用卷积神经网络对火灾信号进行识别的方案。该方案与传统的人工监控、飞机巡视、卫星火情监控等方案对比后认为具有更低的成本以及更高的效率和安全性。
Forest fire is a serious natural disaster and our country is vast, so the forest area is vast, for the prevention and monitoring of forest fire huge demand. Aiming at the problem of forest fire in China, in order to expand the scope of forest fire monitoring and improve the response speed of forest fire identification, a scheme based on UAV image shooting and using convolutional neural network to identify fire signals is designed. Compared with traditional manual monitoring, aircraft patrol, sat-ellite fire monitoring and other programs, the scheme has higher efficiency and safety and lower cost.
出处
《应用数学进展》
2022年第6期3200-3210,共11页
Advances in Applied Mathematics
