Current Advances in Flame Detection using Convolutional Neural Networks

Zhibiao Wang ； Ye Tao ； Jiayi Yu

《International Core Journal of Engineering》 9卷6期 (2023/06) Pp. 65-68

https://doi.org/10.6919/ICJE.202306_9(6).0009

摘要

Flame detection using convolutional neural networks (CNNs) is currently the most advanced and widely used method for detecting fires. Improving CNN performance typically involves increasing recall and precision rates, improving detection speed, and enhancing the overlap between predicted and actual objects. In this paper, we review recent advances in flame detection using CNNs, focusing on precision and recall rate improvements, detection speed optimizations, dataset quality enhancements, and practical applications.

關鍵字

Flame Image Detection ； Convolutional Neural Network ； Recall Rate ； Precision Rate ； Detection Speed ； Prediction Frame ； Actual Object ； Fire Detection

參考文獻

Gaur A, Singh A, Kumar A, et al. Video flame and smoke-based fire detection algorithms: A literature review[J]. Fire Technology, 2020, 56(5): 1943-1980.

Avula S B, Badri S J, Reddy G. A Novel Forest fire detection system using fuzzy entropy optimized thresholding and STN-based CNN[C]//2020 International Conference on Communication Systems & Networks (COMSNETS), Paris, France, 2020: 750-755.

Cao Y, Yang F, Tang Q, et al. An attention enhanced bidirectional LSTM for early forest fire smoke recognition[J]. IEEE Access, 2019, 7:154732-154742.

Shahid M, Hua K. Fire detection using transformer network[C]//2021 Proceedings of the 2021 International Conference on Multimedia Retrieval. 2021: 627-630.

Dutta S, Ghosh S. Forest Fire detection using combined architecture of separable convolution and image processing[C]//2021 1st International Conference on Artificial Intelligence and Data Analytics (CAIDA), Riyadh, Saudi Arabia, 2021: 36-41.