A deep learning approach for early wildfire detection from hyperspectral satellite images

Wildfires are getting more severe and destructive. Due to their fast-spreading nature, wildfires are often detected when already beyond control and consequently cause billion-scale effects in a very short time. Governments are looking for remote sensing methods for early wildfire detection, avoiding billion-dollar losses of damaged properties. The aim of this study was to develop an autonomous and intelligent system built on top of imagery data streams, which is available from around-the-clock satellites, to monitor and prevent fire hazards from becoming disasters. However, satellite data pose unique challenges for image processing techniques, including temporal dependencies across time steps, the complexity of spectral chan-nels, and adversarial conditions such as cloud and illumination. In this paper, we propose a novel wildfire detection method that utilises satellite images in an advanced deep learning architecture for locating wildfires at pixel level. The detection outputs are further visualised in an interactive dashboard that allows wildfire mitigation specialists to deeply analyse regions of interest in the world-map. Our system is built and tested on the Geostationary Operational Environmental Satellites (GOES-16) streaming data source. Empirical evaluations show the superiorperformance of our approach over the baselines with 94% F1- score and 1.5 times faster detections as well as its robustness against different types of wildfires and adversarial conditions.

[1]  Karl Aberer,et al.  An Evaluation of Model-Based Approaches to Sensor Data Compression , 2013, IEEE Transactions on Knowledge and Data Engineering.

[2]  Hong Liu,et al.  A two-step convolutional neural network based computer-aided detection scheme for automatically segmenting adipose tissue volume depicting on CT images , 2017, Comput. Methods Programs Biomed..

[3]  Ali Farhadi,et al.  You Only Look Once: Unified, Real-Time Object Detection , 2015, 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).

[4]  Karl Aberer,et al.  Minimizing Efforts in Validating Crowd Answers , 2015, SIGMOD Conference.

[5]  Yoram J. Kaufman,et al.  A Review of AVHRR-based Active Fire Detection Algorithms: Principles, Limitations, and Recommendations , 2000 .

[6]  G. Roberts,et al.  New GOES imager algorithms for cloud and active fire detection and fire radiative power assessment across North, South and Central America , 2010 .

[7]  Peng Liu,et al.  Ieee Journal of Selected Topics in Applied Earth Observations and Remote Sensing 1 Active Deep Learning for Classification of Hyperspectral Images , 2022 .

[8]  Kai Yang,et al.  A Deep Active Survival Analysis Approach for Precision Treatment Recommendations: Application of Prostate Cancer , 2018, Expert Syst. Appl..

[9]  Andrew K. Skidmore,et al.  Advances in active fire detection using a multi-temporal method for next-generation geostationary satellite data , 2018, Int. J. Digit. Earth.

[10]  Sabira Khatun,et al.  The Classification of EEG Signal Using Different Machine Learning Techniques for BCI Application , 2018, RiTA.

[11]  Matthias Weidlich,et al.  Computing Crowd Consensus with Partial Agreement , 2018, IEEE Transactions on Knowledge and Data Engineering.

[12]  Shanjun Mao,et al.  Spectral–spatial classification of hyperspectral images using deep convolutional neural networks , 2015 .

[13]  David R. Easterling,et al.  Impacts, Risks, and Adaptation in the United States: Fourth National Climate Assessment, Volume II , 2017 .

[14]  Catarina Eloy,et al.  Classification of breast cancer histology images using Convolutional Neural Networks , 2017, PloS one.

[15]  Mourad Ouzzani,et al.  Data Curation with Deep Learning [Vision]: Towards Self Driving Data Curation , 2018, ArXiv.

[16]  Sadasivan Puthusserypady,et al.  A deep learning approach for real-time detection of atrial fibrillation , 2019, Expert Syst. Appl..

[17]  Jun Jo,et al.  Detection and Classification of Vehicle Types from Moving Backgrounds , 2017, RiTA.

[18]  Steven Verstockt,et al.  Review of wildfire detection using social media , 2014 .

[19]  W. Schroeder,et al.  The New VIIRS 375 m active fire detection data product: Algorithm description and initial assessment , 2014 .

[20]  Geoffrey E. Hinton,et al.  ImageNet classification with deep convolutional neural networks , 2012, Commun. ACM.

[21]  Lorenzo Rosasco,et al.  Why and when can deep-but not shallow-networks avoid the curse of dimensionality: A review , 2016, International Journal of Automation and Computing.

[22]  Karl Aberer,et al.  An Evaluation of Aggregation Techniques in Crowdsourcing , 2013, WISE.

[23]  Hyun Myung,et al.  Accurate Localization in Urban Environments Using Fault Detection of GPS and Multi-sensor Fusion , 2015, RiTA.

[24]  Donald J. Wuebbles,et al.  Climate Science Special Report: Fourth National Climate Assessment, Volume I , 2017 .

[25]  C. Justice,et al.  The collection 6 MODIS active fire detection algorithm and fire products , 2016, Remote sensing of environment.

[26]  D. Roberts,et al.  WILDFIRE DETECTION FOR RETRIEVING FIRE TEMPERATURE FROM HYPERSPECTRAL DATA , 2008 .

[27]  Marta C. González,et al.  Using Convolutional Networks and Satellite Imagery to Identify Patterns in Urban Environments at a Large Scale , 2017, KDD.

[28]  Susan Pascoe,et al.  Interim report: 2009 Victorian Bushfires Royal Commission , 2009 .

[29]  Karl Aberer,et al.  Answer validation for generic crowdsourcing tasks with minimal efforts , 2017, The VLDB Journal.

[30]  Taesup Moon,et al.  UDLR Convolutional Network for Adaptive Image Denoiser , 2018, RiTA.

[31]  Siqi Ren,et al.  Multi-object Tracking with Pre-classified Detection , 2017, RiTA.