Forest Fire Detection Method Based on Dual-Branch Multi-Scale Adaptive Feature Fusion Network

Wu, Qinggan; Wei, Chen; Sun, Ning; Xiong, Xiong; Xia, Qingfeng; Zhou, Jianmeng; Feng, Xingyu

doi:10.3390/f16081248

Open AccessArticle

Forest Fire Detection Method Based on Dual-Branch Multi-Scale Adaptive Feature Fusion Network

by

Qinggan Wu

¹,

Chen Wei

²

,

Ning Sun

^1,*

,

Xiong Xiong

²

,

Qingfeng Xia

¹,

Jianmeng Zhou

² and

Xingyu Feng

²

¹

School of Automation, Wuxi University, Wuxi 214105, China

²

Jiangsu Collaborative Innovation Center of Atmospheric Environment and Equipment Technology, Nanjing University of Information Science and Technology, Nanjing 210044, China

^*

Author to whom correspondence should be addressed.

Forests 2025, 16(8), 1248; https://doi.org/10.3390/f16081248

Submission received: 24 June 2025 / Revised: 22 July 2025 / Accepted: 28 July 2025 / Published: 31 July 2025

(This article belongs to the Section Natural Hazards and Risk Management)

Download

Browse Figures

Versions Notes

Abstract

There are significant scale and morphological differences between fire and smoke features in forest fire detection. This paper proposes a detection method based on dual-branch multi-scale adaptive feature fusion network (DMAFNet). In this method, convolutional neural network (CNN) and transformer are used to form a dual-branch backbone network to extract local texture and global context information, respectively. In order to overcome the difference in feature distribution and response scale between the two branches, a feature correction module (FCM) is designed. Through space and channel correction mechanisms, the adaptive alignment of two branch features is realized. The Fusion Feature Module (FFM) is further introduced to fully integrate dual-branch features based on the two-way cross-attention mechanism and effectively suppress redundant information. Finally, the Multi-Scale Fusion Attention Unit (MSFAU) is designed to enhance the multi-scale detection capability of fire targets. Experimental results show that the proposed DMAFNet has significantly improved in mAP (mean average precision) indicators compared with existing mainstream detection methods.

Keywords: forest fire detection; dual branch network; convolutional neural network (CNN); transformer; feature correction; feature fusion

Share and Cite

MDPI and ACS Style

Wu, Q.; Wei, C.; Sun, N.; Xiong, X.; Xia, Q.; Zhou, J.; Feng, X. Forest Fire Detection Method Based on Dual-Branch Multi-Scale Adaptive Feature Fusion Network. Forests 2025, 16, 1248. https://doi.org/10.3390/f16081248

AMA Style

Wu Q, Wei C, Sun N, Xiong X, Xia Q, Zhou J, Feng X. Forest Fire Detection Method Based on Dual-Branch Multi-Scale Adaptive Feature Fusion Network. Forests. 2025; 16(8):1248. https://doi.org/10.3390/f16081248

Chicago/Turabian Style

Wu, Qinggan, Chen Wei, Ning Sun, Xiong Xiong, Qingfeng Xia, Jianmeng Zhou, and Xingyu Feng. 2025. "Forest Fire Detection Method Based on Dual-Branch Multi-Scale Adaptive Feature Fusion Network" Forests 16, no. 8: 1248. https://doi.org/10.3390/f16081248

APA Style

Wu, Q., Wei, C., Sun, N., Xiong, X., Xia, Q., Zhou, J., & Feng, X. (2025). Forest Fire Detection Method Based on Dual-Branch Multi-Scale Adaptive Feature Fusion Network. Forests, 16(8), 1248. https://doi.org/10.3390/f16081248

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Menu

Forest Fire Detection Method Based on Dual-Branch Multi-Scale Adaptive Feature Fusion Network

Abstract

Share and Cite

Article Metrics

Article Access Statistics

Further Information

Guidelines

MDPI Initiatives

Follow MDPI