Abstract:Wildfires are a disastrous phenomenon which cause damage to land, loss of property, air pollution, and even loss of human life. Due to the warmer and drier conditions created by climate change, more severe and uncontrollable wildfires are expected to occur in the coming years. This could lead to a global wildfire crisis and have dire consequences on our planet. Hence, it has become imperative to use technology to help prevent the spread of wildfires. One way to prevent the spread of wildfires before they become too large is to perform early detection i.e, detecting the smoke before the actual fire starts. In this paper, we present our Wildfire Detection and Alert System which use machine learning to detect wildfire smoke with a high degree of accuracy and can send immediate alerts to users. Our technology is currently being used in the USA to monitor data coming in from hundreds of cameras daily. We show that our system has a high true detection rate and a low false detection rate. Our performance evaluation study also shows that on an average our system detects wildfire smoke faster than an actual person.

What problem does this paper attempt to address?

The problem that this paper attempts to solve is **early detection of wildfires**, specifically detecting smoke before a fire occurs through image recognition technology, thereby achieving early warning of wildfires. With the increase in temperature and the intensification of drought conditions caused by climate change, there may be more severe and uncontrollable wildfires in the future. Therefore, it is crucial to use technological means to prevent the spread of wildfires. ### Main problems: 1. **Hazards of wildfires**: Every year, wildfires burn millions of acres of land, destroy thousands of buildings, and cause serious air pollution. In 2021 alone, more than 58,000 fires occurred in the United States, burning 7.13 million acres of land. 2. **Impact of climate change**: Global warming leads to warmer and drier climate conditions, increasing the dryness of soil and vegetation, making these areas more prone to fires. Research shows that in the coming decades, due to the impact of climate change, there will be more large - scale and severe wildfires globally. 3. **Limitations of existing technologies**: Although existing technological tools (such as fire spread simulators, satellite data monitoring, etc.) are helpful for analyzing and mitigating the spread of wildfires, they are insufficient in terms of early detection and prevention. For example, relying on human operators to monitor camera data in real - time is inefficient and error - prone. ### Solutions: The paper proposes an image - based **early wildfire detection system**, which uses machine - learning algorithms to detect smoke from real - time camera data and can send alerts to users in a timely manner. The advantages of this system include: - **High true positive rate**: It can accurately detect wildfire smoke. - **Low false positive rate**: It reduces false alarms. - **Fast response**: The average detection speed is faster than that of human operators. ### Method overview: 1. **Data collection**: 90,000 images were obtained from Alert Wildfire Cameras and labeled. 2. **Model architecture**: ResNet - 18 is used as the backbone network, combined with the Feature Pyramid Network (FPN) and deformable convolution to adapt to smoke detection at different scales. 3. **Training sample selection**: An improved Adaptive Training Sample Selection method (ATSS) is adopted to improve the selection quality of positive samples. 4. **Performance evaluation**: Through the performance in the validation set and practical applications, the efficiency and accuracy of this system are proved. In conclusion, this paper aims to help prevent the spread of wildfires by developing an efficient early wildfire detection system, thereby protecting human communities and the natural environment.

Image-based Early Detection System for Wildfires

An AI-based Early Fire Detection System Utilizing HD Cameras and Real-time Image Analysis

Wildfire Detection Through Deep Learning Based on Himawari-8 Satellites Platform

A Deep Learning-Based Experiment on Forest Wildfire Detection in Machine Vision Course

Wildfire Smoke Detection with Computer Vision

Preliminary Wildfire Detection Using State-of-the-art PTZ (Pan, Tilt, Zoom) Camera Technology and Convolutional Neural Networks

An Efficient Wildfire Detection System for AI-Embedded Applications Using Satellite Imagery

An Intelligent Wildfire Detection Approach through Cameras Based on Deep Learning

Multimodal Wildland Fire Smoke Detection

A Wildfire Smoke Detection System Using Unmanned Aerial Vehicle Images Based on the Optimized YOLOv5

Spot Fire: An Intelligent Forest Fire Detection System Design With Machine Learning

A Multi-Modal Wildfire Prediction and Personalized Early-Warning System Based on a Novel Machine Learning Framework

Wildfire Impact Analysis and Spread Dynamics Estimation on Satellite Images Using Deep Learning

AI-based Wildfire Prevention, Detection and Suppression System

Deep Learning-Based Wildfire Image Detection and Classification Systems for Controlling Biomass

A Survey on IoT Ground Sensing Systems for Early Wildfire Detection: Technologies, Challenges and Opportunities

AI Enabled Fire-Detecting Drone Using Deep Learning Model

Deep Learning Approaches for Wildland Fires Using Satellite Remote Sensing Data: Detection, Mapping, and Prediction

Deep learning approach for uav-based wildfire detection and segmentation

Active Fire Detection Using a Novel Convolutional Neural Network Based on Himawari-8 Satellite Images

Proactive Wildfire Detection and Management using AI, ML, and 5G Technology in the United States