An Edge Computing Environment for Early Wildfire Detection

Abstract

Recently, an increasing demand is growing for installing a rapid response system in forest regions to enable an immediate and appropriate response to wildfires before they spread across vast areas. This paper introduces a multilevel system for early wildfire detection to support public authorities to immediately specify and attend to emergency demands. The presented work is designed and implemented within Edge Computing Infrastructure. At the first level; the dataset samples of wildfire represented by a set of video sequences are collected and labelled for training mode purposes. Then, YOLOv5 deep learning model is adopted in our framework to build a trained model for distinguishing the fire event against non-fire events in binary classification. The proposed system structure comprises IoT entities provided with camera sensor capabilities and NVIDIA Jetson Nano Developer kit as an edge computing environment. At the first level, a video camera is employed to assemble environment information received by the micro-controller middle level to handle and detect the possible fire event presenting in the interested area. The last level is characterized as making a decision by sending a text message and snapshot images to the cloud server. Meanwhile, a set of commands are sent to IoT nodes to operate the speakers and sprinklers, which are strategically assumed to place on the ground to give an alarm and prevent wildlife loss. The proposed system was tested and evaluated using a wildfire dataset constructed by our efforts. The experimental results exhibited 98% accurate detection of fire events in the video sequence. Further, a comparison study is performed in this research to confirm the results obtained from recent methods.

Counting of shoots of Eucalyptus sp. clones with convolutional neural network

“…The processing time required for the DS01 and DS10 datasets was 15.52 and 3.67 min, respectively. Although, according to Mahdi & Mahmood (2022), smaller images contribute to a reduced processing time, the results of the present study may be related to the fact that, when only one shoot per image was considered (DS01), the number of files required was ten times higher than for sampling with ten shoots per image (DS10).…”

“…The relatively short training time was also attributed to the use of Google Colaboratory, which provided graphic processing units (GPUs) for free in its virtual machine in the cloud (Mirhaji et al, 2021), eliminating the need for a high-cost local machine and leading to a significant difference in artificial intelligence applications (Mahdi & Mahmood, 2022). For the present work, a Tesla T4 GPU with 15GB of RAM and 40 processing cores was made available.…”

“…Although there are studies on tree identification using aerial imagery and segmentation (Hentz et al, 2018) or classification algorithms (Ruza et al, 2017;Silva et al, 2018), the task of counting individuals faces challenges similar to those involved in the use of object detection algorithms in computer vision models, mainly convolutional neural networks. These models have been used for various purposes within forest sciences, such as studying the quality of forest roads, wood cracks (Ma et al, 2022), presence of knots in wooden boards (Fang et al, 2021), recognition of leaf diseases (Chen et al, 2022), identification of forest pests (Zhang et al, 2021;Yun et al, 2022), fire detection (Hossain et al, 2020;Lu et al, 2022;Mahdi & Mahmood, 2022;Zhao et al, 2022), and identification of dead trees (Li et al, 2022). All of these studies tested several versions of the You Only Look Once (YOLO) algorithm, originally proposed by Redmon et al (2016) and considered one of the most powerful algorithms for object detection.…”

Design of a prototype for sending fire notifications in homes using fuzzy logic and internet of things

“…The open system for fire detection is used in forested areas to reduce wildlife losses using system training, improving the efficiency of the system by using cameras connected to a server to identify potential fire situations. In addition, fuzzy logic determines the degree of assertiveness to classify it as a possible fire and trigger the actuators installed in the area [25]. Advances in the sensitivity and availability of gas detection sensors and other parameters have been significant in recent years due to the use of technology and research into their functionalities, which in time have been decisive in the timely detection of potential fires in houses, buildings, and other environments [22].…”

Facial recognition based on enhanced neural network

“…The quick progress of machine learning-based systems has made HCI applicable to many new domains, including unmanned aerial vehicles (UAV) detection [15], Lung segmentation [16], speaker identification [17], video summarization [18], handwriting recognition [19], contextual anomaly detection [20], temperature prediction [21], food recognition [22], face retrieval system [23], wildfire detection [24], and many more. Therefore, an unsupervised neural network that contrasts favorably with other techniques has been used.…”