Forest fire pre-warning method and system based on deep learning

Abstract

The invention discloses a forest fire pre-warning method and system based on deep learning and belongs to the field of firefighting safety. The method comprises: S1, allowing an unmanned aerial vehicle to sense, in real time, temperature and humidity information of a driving area, taking a forest image of the driving area, transmitting the forest image to a ground station; synchronously transmitting locational information to the ground station; S2, judging fire pre-warning according to the temperature and humidity information, and transmitting a fire pre-warning signal to the ground station; S3, allowing the ground station to receive the fire pre-warning signal, and using a deep learning algorithm to process the forest image to obtain whether a fire is about to occur or whether fire results are present; S4, transmitting fire information to a forest management center. The temperature and humidity information is sensed and subjected to comparison; visual judgment is made on chances for afire to occur; if chances for the fire to occur are great, the forest image is processed via the deep learning algorithm; fire pre-occurrence and post-occurrence conditions can be precisely recognized; it is convenient for relevant personnel to accurately acquire fire conditions in the first place.

Description

technical field [0001] The invention relates to the field of fire safety, in particular to a forest fire early warning method and system based on deep learning. Background technique [0002] The definition of forest fire is: in the forest area, the sudden burning of large forest trees that are out of human control, and the spread speed is very fast. Forest fire prevention is an important part of China's disaster prevention and reduction. It is of great significance to the protection of forest resources and the development of a good ecological environment, and has a major impact on China's energy development. [0003] Forest fire prevention monitoring mainly adopts the methods of artificial gaze, remote video surveillance, satellite remote sensing and drone patrol. The artificial watch method is to set up watch posts at the commanding heights, and the on-duty personnel are on duty 24 hours a day. Due to human negligence and negligence, many fires will not be detected early, ...

AI Technical Summary

Problems solved by technology

Since the infrared thermal imager relies on temperature difference imaging, and the temperature difference of the general target is not large, the contrast of the infrared thermal image is low, which makes the ability to distinguish details worse, and the target cannot be seen clearly through transparent obstacles; while the commonly used camera video image processing method It is impossible to accurately identify the pre-fire situation and the fire situation

Images