Fire Ignition Library (FIgLib)
收藏资源简介:
Fire Ignition Library (FIgLib)是由加州大学圣地亚哥分校创建的一个公开可用数据集,包含近25,000张标记的野火烟雾图像,这些图像来自南加州偏远山顶上部署的固定视角摄像头。数据集反映了从2016年6月至2021年7月期间,315次火灾序列的图像,每序列通常包含火灾前后40分钟的图像,间隔约60秒。FIgLib数据集旨在为野火烟雾检测提供一个大规模、标记公开的数据集,以支持深度学习方法的研究,特别是用于实时野火烟雾检测,从而实现自动化通知系统,减少野火响应时间。
Fire Ignition Library (FIgLib) is a publicly available dataset developed by the University of California, San Diego. It contains nearly 25,000 labeled wildfire smoke images captured by fixed-view cameras deployed on remote mountain tops in Southern California. The dataset encompasses images from 315 fire sequences collected between June 2016 and July 2021, with each sequence typically including 40 minutes of imagery spanning before and after the fire event, captured at approximately 60-second intervals. The primary objective of FIgLib is to offer a large-scale, publicly accessible labeled dataset for wildfire smoke detection research, specifically to support deep learning-based methods for real-time wildfire smoke detection, thereby enabling automated notification systems to reduce wildfire response times.

- 1FIgLib & SmokeyNet: Dataset and Deep Learning Model for Real-Time Wildland Fire Smoke Detection加州大学圣地亚哥分校 · 2022年



