Abstract
Recently, many researchers have attempted to use convolutional neural networks (CNNs) for wildfire smoke detection. However, the application of CNNs in wildfire smoke detection still faces several issues, e.g., the high false-alarm rate of detection and the imbalance of training data. To address these issues, we propose a novel framework integrating conventional methods into CNN for wildfire smoke detection, which consisted of a candidate smoke region segmentation strategy and an advanced network architecture, namely wildfire smoke dilated DenseNet (WSDD-Net). Candidate smoke region segmentation removed the complex backgrounds of the wildfire smoke images. The proposed WSDD-Net achieved multi-scale feature extraction by combining dilated convolutions with dense block. In order to solve the problem of the dataset imbalance, an improved cross entropy loss function, namely balanced cross entropy (BCE), was used instead of the original cross entropy loss function in the training process. The proposed WSDD-Net was evaluated according to two smoke datasets, i.e., WS and Yuan, and achieved a high AR (99.20%) and a low FAR (0.24%). The experimental results demonstrated that the proposed framework had better detection capabilities under different negative sample interferences.
Highlights
Wildfires destroy the natural ecological environment, and threaten human safety and property [1]
Since image-based fire detection effectively reduces outside interference compared to the currently available sensors, image-based fire detection has become a hot topic in modern wildfire alarm systems [2]
To address the above challenges, we propose a novel framework integrating conventional methods into convolutional neural networks (CNNs) for wildfire smoke detection, consisting of a candidate smoke region segmentation strategy and advanced network architecture, namely wildfire smoke dilated DenseNet (WSDD-Net)
Summary
Wildfires destroy the natural ecological environment, and threaten human safety and property [1]. Since image-based fire detection effectively reduces outside interference compared to the currently available sensors, image-based fire detection has become a hot topic in modern wildfire alarm systems [2]. Fire is often accompanied by smoke, which is emitted faster than flames. Smoke detection is an effective way to recognize potential fire disasters at the beginning of a breakout
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have