Recent Advances in Fire Detection and Monitoring Systems: A Review

Wildfires are one of the most impacting natural disasters, leading to a huge devastation of humans and the environment. Due to the rapid development of sensors and technologies as well as the success of computer vision algorithms new and complete solutions for automatic fire monitoring and detection have been exposed. However, in the past years, only few literature reviews have been proposed to cover researches until the year 2015. To fill this gap, we provide, in this paper, an up-to-date comprehensive review on this problem. First, we present a general description and a comparative analysis in terms of reliability, flexibility and efficiency, of these systems. Then, we expose vision-based methods for fire detection. Our main focus was on techniques based on deep convolutional neural networks (CNNs).