Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2026 Jan 20;21(1):e0338794.
doi: 10.1371/journal.pone.0338794. eCollection 2026.

Forest fire prediction using image processing

Yingdan Li  1   2 , Junting Chen  1 , Yaxuan Zeng  1 , Yuanyuan Ding  1 , Chaobing Huang  3 , Hongxing Tian  1
Affiliations

Forest fire prediction using image processing

Yingdan Li et al. PLoS One. .

Abstract

Forest fires pose a significant threat to public safety and the environment, and harmful pollutants spread rapidly in areas covered by vegetation. Early detection is very important for preventing forest fires from evolving into catastrophic fires. The traditional prediction methods have relatively low accuracy. They can only identify fires clearly after they occur, making it difficult to meet the requirements of precise real-time detection. The YOLOv5-PSG model proposed in this paper improves the YOLOv5 model. After 300 rounds of training, the average recognition accuracy rate of mAP can reach 93.1%, and the accuracy rate can reach approximately 0.802. After 300 rounds of training and learning, the confidence level can reach about 0.965. This improvement makes fire early warning and prediction more comprehensive and effective, ultimately protecting human life and the environment by mitigating the impact of wildfires.

PubMed Disclaimer

Conflict of interest statement

The authors have declared that no competing interests exist.

Figures

Fig 1
Fig 1. YOLOv5-PSG network structure diagram.
Fig 2
Fig 2. System design drawing.
Fig 3
Fig 3. Fire source smoke dataset.
Fig 4
Fig 4. Candlelight dataset.
Fig 5
Fig 5. Target annotation page.
Fig 6
Fig 6. Visual comparison of candlelight.
Fig 7
Fig 7. Visual comparison of high-altitude forest fire sources in complex scenes.
Fig 8
Fig 8. Comparison of high-altitude smoke visualisations.
Fig 9
Fig 9. Diagram of data enhancement using arbitrary scaling cuts [20].
Fig 10
Fig 10. Diagram of positioning picture categories [20].
Fig 11
Fig 11. Confidence graph for test pictures [20].
Fig 12
Fig 12. Labels diagram.
Fig 13
Fig 13. Precision_curve diagram.
Fig 14
Fig 14. Labels.correlogram.
See this image and copyright information in PMC

References

    1. Liu Y . Climate Change Science Dynamic Monitoring Bulletin. Lanzhou Literature and Information Center, Chinese Academy of Sciences. 2024.
    1. Li C. Application analysis of target detection technology in “intelligent fire protection”. China Fire Protection. 2023;S1:66–7, 70.
    1. Zhang J, Peng D, Zhang C, et al. Forest fire prediction modeling based on deep learning in the Greater Hinggan Mountains of Inner Mongolia. Research in Forestry Sciences. 2024;37(01):31–40.
    1. Deepa KR, Chaitra AS, Jhansi K, Anitha Kumari RD, Mallikarjun MK. Development of fire detection surveillance using machine learning & IoT. In: 2022 IEEE 2nd Mysore Sub Section International Conference (MysuruCon), 2022. 1–6.
    1. Ganesan V, Ramasamy V, Manoj C, Tejaswi T. Contextual Emotional Classifier: An Advanced AI-Powered Emotional Health Ecosystem for Women Utilizing Edge Devices. TS. 2023;40(6):2481–94. doi: 10.18280/ts.400613 - DOI