Flame detection method and device

Abstract

The embodiment of the invention discloses a flame detection method. The method comprises the following steps: acquiring a detection picture; inputting a plurality of detection pictures into a flame detection network to obtain suspected flame areas surrounded by candidate frames on the detection pictures; cutting the detection pictures according to a candidate box, and deforming the detection pictures into suspected flame pictures with the same size; wherein the flame detection network is a two-stage cascaded convolutional network in which a self-varying convolutional layer is added to a candidate region extraction layer; inputting a suspected flame picture into a 3D convolutional network, and extracting spatial features and low-order dynamic features of the suspected flame picture; and inputting the spatial features and the low-order dynamic features into a recurrent neural network, extracting the high-order dynamic features of the suspected flame picture, and judging whether the suspected flame picture has flames or not. Through the flame detection method provided by the embodiment of the invention, flames in the detection picture can be identified more accurately and quickly.

Description

technical field [0001] The invention relates to the field of image detection, in particular to a method and device for detecting flames. Background technique [0002] With the rapid development of the economy, various high-level comprehensive construction groups continue to emerge. In high-rise buildings, due to dense population and concentrated property, the fire protection problem is also particularly prominent. Although there are mature flame detection methods now, such as smoke detectors, temperature detectors, and photosensitive detectors, the flames are detected using the characteristics of smoke, temperature, and light of flames. [0003] However, in comprehensive shopping malls, due to the large number of merchants, the back kitchens of catering merchants are relatively small and closed, and it is impossible to install smoke, temperature, and photodetectors. The intensity of the flame detection is still not up to the alarm levels of smoke, temperature, and photosen...

AI Technical Summary

Problems solved by technology

[0003] However, in comprehensive shopping malls, due to the large number of merchants, the back kitchens of catering merchants are relatively small and closed, and it is impossible to install smoke, temperature, and photodetectors. The intensity of the fire is still not up to the alarm level of smoke, temperature, and photosensitive detectors, which brings a certain degree of difficulty to flame detection.

As a result, the fire in the shopping mall is a big fire when it is discovered, and it is extremely difficult to extinguish the work, and the flow of people in the shopping mall is relatively large. Once a fire occurs, it is bound to cause greater casualties

Images