Unmanned aerial vehicle forest fire detection system

Abstract

The invention discloses an unmanned aerial vehicle forest fire detection system, which comprises an unmanned aerial vehicle, an image processing unit, a control unit and a fire alarm, wherein the unmanned aerial vehicle shoots an image of a forest site and transmits the image to the image processing unit; the image processing unit performs initial fire target detection on the image and sends a detection result to the control unit; and the control unit judges whether a fire occurs according to the detection result, if the fire occurs, a starting instruction is sent to the fire alarm, and the fire alarm gives an alarm according to the starting instruction. Accurate detection of initial fires is realized.

Description

technical field [0001] The invention relates to an unmanned aerial vehicle forest fire detection system. Background technique [0002] Domestic forest fire monitoring methods can be divided into four levels: satellite monitoring, aerial monitoring, near-ground observation and ground patrol according to spatial location, and a three-dimensional forest fire monitoring system has basically been formed. Satellite monitoring has the advantages of high timeliness, wide range, and strong continuity. However, current satellites do not have the ability to penetrate thick clouds to detect high-temperature thermal radiation under clouds, and satellite monitoring of forest fires is easily blocked by clouds and fog and affected by satellite orbits. influences. Aviation monitoring mainly monitors fires through aviation ranger aircraft or unmanned aircraft, which has a fast response speed, but the supply of aircraft is tight and the cost is extremely high. However, near-surface observa...

AI Technical Summary

Problems solved by technology

Satellite monitoring has the advantages of high timeliness, wide range, and strong continuity. However, current satellites do not have the ability to penetrate thick clouds to detect high-temperature thermal radiation under clouds, and satellite monitoring of forest fires is easily blocked by clouds and fog and affected by satellite orbits. influences

Aviation monitoring mainly monitors the fire situation through aviation ranger aircraft or unmanned aircraft, which has a fast response speed, but the supply of aircraft is tight and the cost is extremely high

However, the near-surface observation and ground patrol observation range is limited, the workload is heavy, and the efficiency is low.

Traditional image processing fire detection methods require manual feature extraction, and manual feature extraction is too dependent on manpower, and has problems such as low recognition accuracy, slow detection speed, high production cost, and easy to be affected by the surrounding environment.

Images