Early forest flame and smoke detection based on improved feature extraction module with enhanced image processing inspired by YOLOV7

Ruipeng Han; Yunfei Liu; Xueyi Kong; Zhihui Qiu; Shuang Li; Han Liu

doi:10.1071/WF24050

RESEARCH ARTICLE (Open Access)

Previous Next Contents Vol 33(12)

Early forest flame and smoke detection based on improved feature extraction module with enhanced image processing inspired by YOLOV7

Ruipeng Han ^A , Yunfei Liu ^A ^* , Xueyi Kong ^A , Zhihui Qiu ^A , Shuang Li ^A and Han Liu ^B

+ Author Affiliations

- Author Affiliations

^A College of Information Science and Technology, Nanjing Forestry University, Nanjing 210037, China. Email: hrp@njfu.edu.cn, xy18956762687@163.com, qiuzhihui@njfu.edu.cn, lishuang@njfu.edu.cn

^B College of Letters and Science, University of Wisconsin-Madison, Madison, WI 53706, USA. Email: hliu568@wisc.edu

^* Correspondence to: lyf@njfu.com.cn

International Journal of Wildland Fire 33, WF24050 https://doi.org/10.1071/WF24050

Submitted: 15 March 2024 Accepted: 29 October 2024 Published: 3 December 2024

© 2024 The Author(s) (or their employer(s)). Published by CSIRO Publishing on behalf of IAWF. This is an open access article distributed under the Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License (CC BY-NC-ND)

Abstract

Background

The forest environment is intricate and dynamic, with its depiction influenced by various factors such as geographical location, weather conditions, and capture angles. Relying solely on flame or smoke is insufficient for precise fire information.

Aims

This paper proposes a method for accurate detection on forest flame and smoke based on improved feature extraction module with enhanced image processing.

Methods

We a fusion-guided filtering image processing method and flame segmentation strategy to augment the quality of dataset. Additionally, an outstanding extraction backbone, incorporating ghost modules and decoupled fully connected (DFC) attention modules, is developed to increase the model’s receptive field. Furthermore, the ELAN-S neck with SimAM attention mechanism is introduced to fuse features from the backbone network, facilitating the extraction of shallow and deep-level semantic information.

Key results

Compared to YOLOV7, our model demonstrates superior performance with a 5% increase in mean average precision (mAP), a 4.3% increase in average precision for small objects (APS), and a 3–4% enhancement in other metrics.

Conclusions

The proposed model achieves a good balance between detection speed and detection accuracy. The improved model performs well in real forest fire detection scenarios.

Implications

In the early forest fire detection, the model considers both flame and smoke information to describe the fire situation, and effectively combines the semantic information of both for fire warning.

Keywords: DFC attention module, early forest fire detection, feature extraction, fire warning, fusion-guided filtering, image processing, semantic information, SimAM attention mechanism.

References

Appana DK, Islam R, Khan SA, et al. (2017) A video-based smoke detection using smoke flow pattern and spatial-temporal energy analyses for alarm systems. Information Sciences 418, 91-101.
| Crossref | Google Scholar |

Cazzolato MT, Avalhais L, Chino D, Ramos JS, de Souza JA, Rodrigues Jr JF, Traina A (2017) Fismo: A compilation of datasets from emergency situations for fire and smoke analysis. In ‘Proceedings of the Brazilian Symposium on Databases-SBBD’, 2–5 October 2017, Uberlândia, Brazil. pp. 213–223.

Chen SJ, Hovde DC, Peterson KA, Marshall AW (2007) Fire detection using smoke and gas sensors. Fire Safety Journal 42, 507-515.
| Crossref | Google Scholar |

Chino DYT, Avalhais LPS, Rodrigues JF and Traina AJM (2015-October) Bowfire: Detection of fire in still images by integrating pixel color and texture analysis. In ‘Brazilian Symposium of Computer Graphic and Image Processing’. 10.1109/SIBGRAPI.2015.19

Dios JM-D, Arrue B, Ollero A, Merino L, Gómez-Rodríguez F (2008) Computer vision techniques for forest fire perception. Image and Vision Computing 26, 550-562.
| Crossref | Google Scholar |

Eugenio FC, dos Santos AR, Fiedler NC, Ribeiro GA, da Silva AG, dos Santos ÁB, Paneto GG, Schettino VR (2016) Applying GIS to develop a model for forest fire risk: a case study in Espírito Santo, Brazil. Journal of Environmental Management 173, 65-71.
| Crossref | Google Scholar | PubMed |

Fernandes AM, Utkin AB, Lavrov AV, Vilar RM (2004) Development of neural network committee machines for automatic forest fire detection using lidar. Pattern Recognition 37, 2039-2047.
| Crossref | Google Scholar |

Frizzi S, Kaabi R, Bouchouicha M, Ginoux JM, Moreau E, Fnaiech F (2016) Convolutional neural network for video fire and smoke detection. In ‘Proceedings of the IECON 2016—42nd Annual Conference of the IEEE Industrial Electronics Society’, 23–26 October 2016, Florence, Italy, pp. 877–882. 10.1109/IECON.2016.7793196

Hossain FA, Zhang YM, Tonima MA (2020) Forest fire flame and smoke detection from UAV-captured images using fire-specific color features and multi-color space local binary pattern. Journal of Unmanned Vehicle Systems 8, 285-309.
| Crossref | Google Scholar |

HPWREN Fire (2020) HPWREN Fire. Available at https://www.hpwren.ucsd.edu/FIgLib/ [accessed 18 August 2020]

Kucuk G, Kosucu B, Yavas A, Baydere S (2008) FireSense: Forest Fire Prediction and Detection System using Wireless Sensor Networks. In ‘Proceedings of the 4th IEEE/ACM International Conference Distributed Computing in Sensor Systems’, 11–14 June 2008, Santorini Island, Greece, pp. 1–4.

Lee B, Kwon O, Jung C, Park S (2001) The development of UV/IR combination flame detector. Journal of KIIS 16, 1-8.
| Google Scholar |

Lu P, Zhao Y, Xu Y (2021) A two-stream CNN Model with adaptive adjustment of receptive field dedicated to flame region detection. Symmetry 13, 397.
| Crossref | Google Scholar |

Noda H, Niimi M (2007) Colorization in YCbCr color space and its application to JPEG images. Pattern Recognition 40(12), 3714-3720.
| Crossref | Google Scholar |

Pan J, Ou X, Xu L (2021) A collaborative region detection and grading framework for forest fire smoke using weakly supervised fine segmentation and lightweight faster-RCNN. Forests 12, 768.
| Crossref | Google Scholar |

Ryu J, Kwak D (2021) Flame detection using appearance-based pre-processing and convolutional neural network. Applied Sciences 11, 5138.
| Crossref | Google Scholar |

Töreyin BU, Dedeoğlu Y, Güdükbay U, Cetin AE (2006) Computer vision based method for real-time fire and flame detection. Pattern Recognition Letters 27, 49-58.
| Crossref | Google Scholar |

VisiFire:COMPUTER VISION BASED FIRE DETECTION SOFTWARE (2021) Available at http://signal.ee.bilkent.edu.tr/VisiFire/ [accessed 1 January 2021]

Yang Z, Bu L, Wang T, et al. (2020) Indoor video flame detection based on lightweight convolutional neural network. Pattern Recognition and Image Analysis 30, 551-564.
| Crossref | Google Scholar |

Ye S, Bai Z, Chen H, et al. (2017) An effective algorithm to detect both smoke and flame using color and wavelet analysis. Pattern Recognition and Image Analysis 27(1), 131-138.
| Crossref | Google Scholar |

Yu L, Wang N, Meng X (2005) Real-time forest fire detection with wireless sensor networks. In ‘Proceedings. 2005 International Conference on Wireless Communications, Networking and Mobile Computing’. Wuhan, China. pp. 1214–1217. 10.1109/WCNM.2005.1544272

Yuan C, Zhang Y, Liu Z (2015) A survey on technologies for automatic forest fire monitoring, detection, and fighting using unmanned aerial vehicles and remote sensing techniques. Canadian Journal of Forest Research 45, 783-792.
| Crossref | Google Scholar |