Abstract
The widespread use of renewable energy resources requires more immediate and effective fire alarms as a preventive measure. The fire is usually weak in the initial stages, which is not conducive to detection and identification. This paper validates a solution to resolve that problem by a flame detection algorithm that is more sensitive to small flames. Based on Yolov3, the parallel convolution structure of Inception is used to obtain multi-size image information. In addition, the receptive field of the convolution kernel is increased with the dilated convolution so that each convolution output contains a range of information to avoid information omission of tiny flames. The model accuracy has improved by introducing a Feature Pyramid Network in the feature extraction stage that has enhanced the feature fusion capability of the model. At the same time, a flame detection database for early fire has been established, which contains more than 30 fire scenarios and is suitable for flame detection under various challenging scenes. Experiments validate the proposed method not only improves the performance of the original algorithm but are also advantageous in comparison with other state-of-the-art object detection networks, and its false positives rate reaches 1.2% in the test set.
Highlights
Yong-Jin Kim (Young-Jin and Eun-Gyung, 2017)tries to apply Faster RCNN to flame detection, and Shen (Shen et al, 2018)simplifies the Yolo (You Only Look Once) network to carry out flame detection, both of which achieve good results, indicating that the flame detection method based on deep learning is superior to the traditional video fire detection method in performance
To test the detection performance of the model, we introduced the following indicators: Precision Rate (PR), Recall Rate (RR), Accuracy Rate (AR), and False Alarm Rate (FAR)
It can be seen that compared with Yolov3, the proposed method has improved in all four indicators, including a significant increase in Precision Rate and a significant decrease in False Alarm Rate, reflecting that the improved model has improved the performance of detecting small flames
Summary
Renewable energy sources is playing an increasingly important role in industry (Qazi et al, 2019). Vision-based fire detection systems can play a decisive role in the flame detection of renewable energy sources. The traditional video fire detection method is based on the characteristics of the flame. Yong-Jin Kim (Young-Jin and Eun-Gyung, 2017)tries to apply Faster RCNN to flame detection, and Shen (Shen et al, 2018)simplifies the Yolo (You Only Look Once) network to carry out flame detection, both of which achieve good results, indicating that the flame detection method based on deep learning is superior to the traditional video fire detection method in performance. The early flame of fire is weak, so it is easy to be ignored by the detection model To solve this problem, this paper proposes a fire detection and identification method based on improved Yolov. This paper has established a flame database for early fires, which involves a variety of fire scenarios to establish a foundation for future flame detection research
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
