Intelligent Airground Collaborative Firefighting System Based on Visual SLAM
With the rapid progress of artificial intelligence technology,drones have been widely used in modern production and daily life in various industries..The use of drones for aerial per-ception is a powerful supplement to traditional ground robot perception.Drones are high-speed,flexible,and have a wide range of applications.They can quickly obtain maps and identify tar-gets.Unmanned vehicles have long endurance,strong load capacity,and can carry various large e-quipment for long-term operations.Combining the advantages of both,the fire source is identi-fied through autonomous patrol by drones,and the unmanned vehicle receives the location of the fire source and autonomously navigates to the target location to extinguish the fire.This design is based on a quadcopter drone platform,using a binocular camera T265 for precise spatial posi-tioning,using the OPENCV visual library to accurately identify the red fire source,sending fire extinguishing instructions and fire coordinates to the unmanned vehicle through wireless com-munication,autonomous path planning,avoiding obstacles to reach the fire location,identifying and extinguishing the fire source,and returning autonomously.The experimental results show that using drone assisted positioning significantly improves speed and accuracy compared to un-manned vehicle exploration alone.The navigation,obstacle avoidance,and perception capabilities of unmanned vehicles have been significantly improved,and this system has good reference sig-nificance for autonomous unmanned firefighting systems.
NETL
NSTL
万方数据
