Description
Fire detection through normal RGB videos for forest fire prevention.
Abstract
Forest fires are an important threat to natural ecosystems, and early detection is essential to prevent damage. Conventional fire detection relies on electronic sensors or watchtowers. These approaches have limitations in outdoor environments and forested areas.
We combine video feeds with a motion detection algorithm and a gradient-boosting model to predict forest fires in real-time. This approach is more cost-effective.
Implementation
Files
Technical Report (3.2 MB)
Project Slides (10.2 MB)
Additional Information
This project was made as part of the 2023 AI Lab: Computer Vision and NLP course at Sapienza Università di Roma.