Fire detecting system applied to unmanned helicopter and fire detecting method thereof

Abstract

The invention discloses a fire detecting system applied to an unmanned helicopter and a fire detecting method thereof. The system comprises infrared image acquiring equipment and a signal processing flow network, wherein the signal processing flow network comprises an input layer, a middle layer and an output layer; the temperature gradient operator calculation is respectively carried out in the vertical direction, the horizontal direction, the direction of plus 45 degrees and the direction of minus 45 degrees in the center of a suspected area, and a calculation result is a basis of judging fire flame; a differential algorithm adopting interframe time difference as the time change rate is carried out by a method for subtracting absolute values of adjacent interframe pixel gray values, anda calculation result is used as a basis of judging a fire high-risk point. The invention rapidly recognizes the fire flame and the fire high-risk point by a platform with low cost and high efficiencyof the unmanned helicopter through an infrared camera, early warns a fire accurately in time and has the characteristics of manoeuvrability, high efficiency and no dead angle. A fire flame and fire high-risk point detecting algorithm designed by aiming at the detection of an infrared image is efficient, simple, practical and effective.

Description

technical field [0001] The invention relates to a fire detection system, in particular to a fire detection system applied to a drone and a fire detection method thereof. Background technique [0002] Forest fire is one of the major disasters that cause huge losses to people's lives and properties. The detection and early warning of forest fires has always been an important part of fire prevention. At present, forest fire prevention is divided into the following categories: 1. Manual inspection; 2. Fixed-position pyrotechnic sensors and cameras; 3. Use various civil aircraft to patrol and check fire points. [0003] Manual mountain inspection is cheap and is widely used as a forest fire prevention method. However, the ability to identify with the naked eye is extremely limited. It has no ability to identify early fires that have not yet formed flames, and cannot quickly identify early fires. [0004] The smoke sensor installed in a fixed position can effectively cover a larg...

AI Technical Summary

Problems solved by technology

However, due to the terrain constraints of forests and mountains, the installation position of the pyrotechnic sensor is limited to a certain extent, so the possibility of blind spots for observation cannot be avoided

In addition, to obtain an effective fire detection effect, a large number of sensors are required, and the high maintenance costs of these sensors also increase the cost of fire prevention.

[0005] Various civil aircraft patrol inspections can effectively and evenly detect forest fires, and there is no blind spot for observation. However, the rental cost of civil aircraft is extremely high, resulting in high fire prevention costs, and the inspection and inspection capabilities for high-risk fire points with coverings limited

[0006] With the development of drone technology, drones are playing an increasingly important role in many fields, but there is still a gap in drone fire detection using infrared cameras

In addition, in the field of fire detection algorithms, there are flame detection algorithms based on color images, smoke and fire detection algorithms, brightness detection algorithms, etc., which are relatively mature flame detection algorithms, but these flame detection algorithms are only suitable for fire detection of color images. However, the ideal detection effect cannot be achieved in the fire detection of infrared images.

Images