Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Deep Learning Based Wildfire Detection for Peatland Fires Using Transfer Learning

Emadeldeen Hamdan, Ahmad Faiz Tharima, Mohd Zahirasri Mohd Tohir, Dayang Nur Sakinah Musa, Erdem Koyuncu, Adam J. Watts, Ahmet Enis Cetin

TL;DR

Experimental results demonstrate that transfer learning significantly improves detection accuracy and robustness compared to training from scratch, particularly under challenging conditions such as low-contrast smoke, partial occlusions, and variable illumination.

Abstract

Machine learning (ML)-based wildfire detection methods have been developed in recent years, primarily using deep learning (DL) models trained on large collections of wildfire images and videos. However, peatland fires exhibit distinct visual and physical characteristics -- such as smoldering combustion, low flame intensity, persistent smoke, and subsurface burning -- that limit the effectiveness of conventional wildfire detectors trained on open-flame forest fires. In this work, we present a transfer learning-based approach for peatland fire detection that leverages knowledge learned from general wildfire imagery and adapts it to the peatland fire domain. We initialize a DL-based peatland fire detector using pretrained weights from a conventional wildfire detection model and subsequently fine-tune the network using a dataset composed of Malaysian peatland images and videos. This strategy enables effective learning despite the limited availability of labeled peatland fire data. Experimental results demonstrate that transfer learning significantly improves detection accuracy and robustness compared to training from scratch, particularly under challenging conditions such as low-contrast smoke, partial occlusions, and variable illumination. The proposed approach provides a practical and scalable solution for early peatland fire detection and has the potential to support real-time monitoring systems for fire prevention and environmental protection.

Deep Learning Based Wildfire Detection for Peatland Fires Using Transfer Learning

TL;DR

Experimental results demonstrate that transfer learning significantly improves detection accuracy and robustness compared to training from scratch, particularly under challenging conditions such as low-contrast smoke, partial occlusions, and variable illumination.

Abstract

Machine learning (ML)-based wildfire detection methods have been developed in recent years, primarily using deep learning (DL) models trained on large collections of wildfire images and videos. However, peatland fires exhibit distinct visual and physical characteristics -- such as smoldering combustion, low flame intensity, persistent smoke, and subsurface burning -- that limit the effectiveness of conventional wildfire detectors trained on open-flame forest fires. In this work, we present a transfer learning-based approach for peatland fire detection that leverages knowledge learned from general wildfire imagery and adapts it to the peatland fire domain. We initialize a DL-based peatland fire detector using pretrained weights from a conventional wildfire detection model and subsequently fine-tune the network using a dataset composed of Malaysian peatland images and videos. This strategy enables effective learning despite the limited availability of labeled peatland fire data. Experimental results demonstrate that transfer learning significantly improves detection accuracy and robustness compared to training from scratch, particularly under challenging conditions such as low-contrast smoke, partial occlusions, and variable illumination. The proposed approach provides a practical and scalable solution for early peatland fire detection and has the potential to support real-time monitoring systems for fire prevention and environmental protection.
Paper Structure (7 sections, 2 equations, 2 figures, 1 table)

This paper contains 7 sections, 2 equations, 2 figures, 1 table.

Table of Contents

  1. Introduction
  2. Backgroud
  3. Transfer Learning
  4. Data PreProcessing
  5. Hadamard Transform based Deep Neural Network
  6. Results
  7. Conclusion

Figures (2)

  • Figure 1: Peatland Wildfire Smoke Detection Examples: Two examples of peatland wildfire detection: (A) and (B). The image frame is partitioned into overlapping $224\times224$ patches. Green boxes indicate no fire/smoke, while red boxes indicate fire/smoke. Numbers inside the boxes denote the predicted fire probability for each patch. Boxes without probabilities or with partial borders occur near frame edges due to the overlapping window scheme.
  • Figure 2: Walsh--Hadamard Transform (WHT) layer architecture. WHT and IWHT denote the forward and inverse Walsh--Hadamard transforms, respectively. The scaling layer performs element--wise multiplication with $N$ trainable parameters, contains no bias term, and uses a hard--thresholding activation.