potential of deep learning in satellite-based fire detection and ... · nceo king’s college...
TRANSCRIPT
Tianran Zhang, Martin Wooster, Daniel Fisher
NCEO King’s College London
Potential of Deep Learning in Satellite-based Fire Detection and Emission Estimation Study
2
Tianran Zhang | [email protected] | NCEO King's College London
Aqua-MODIS15 Sept 2014
Sept 2014
Jan 2015
April 2014
Copernicus Atmosphere Monitoring
Service (CAMS)
MODIS FRP
Based
Global Fire Assimilation System (GFAS)
3
Tianran Zhang | [email protected] | NCEO King's College London
VIIRS Channels
Potential Application 1:
Fire Detection Alg. Using Multi-spectral Remote Sensing Data
Input: 28x28 window data for all 16 VIIRS-M band channels
Output: Cloud/water/land/fire Mask
4
Tianran Zhang | [email protected] | NCEO King's College London
Potential Application 1:
Fire Detection Alg. Using Multi-spectral Remote Sensing Data
Results Analysis: Overfitting 5
Tianran Zhang | [email protected] | NCEO King's College London
Data Scientist Solution
• More training data.
• Change model structure.
• Apply regularization.
Illustration of overfitting.
(source: Wikipedia). Remote Sensing Scientist Solution
• Look into data details (geometry, seasonal/regional special pattern).
• Improve data quality by using other source/higher resolution data.
Results Analysis 6
Tianran Zhang | [email protected] | NCEO King's College London
t-Distributed Stochastic NeighborEmbeddingNonlinear dimensionality reduction technique for embedding high-dimensional data for visualization in a low-dimensional space.
1
2 3
4 5
Potential Application 2:
Real Time Multi Masks Generation 7
Tianran Zhang | [email protected] | NCEO King's College London
8
Tianran Zhang | [email protected] | NCEO King's College London
MODIS VIIRS
VIIRS 26/07/2019
Current fire products suffer 10-40% false fire detection in certain urban area of Eastern China.
Potential Application 3:
Improving Accuracy of Fire Detection Alg.
9
Tianran Zhang | [email protected] | NCEO King's College London
GlobeLand30 land cover product (Chen et al, 2015) re-gridded to 0.01degree in eastern China (111-123° E, 27-40° N).
• Landcover map product has long delay.
• Data not available for real-time fire detection.
• Dynamic world (urbanisation).• Spatial & temporal resolution.(Seasonal fire patterns vs. yearly map).
Potential Application 3:
Improving Accuracy of Fire Detection Alg.
10
Tianran Zhang | [email protected] | NCEO King's College London
Source: University of Cincinnati
VIIRS SDR data(L1B product)Training Stage Labels:
VIIRS Active Fire Product + Landcover MapVegetation MapCultivation Map……
Deploying StageVIIRS SDR data(L1B product)
Results
Potential Application 3:
Improving Accuracy of Fire Detection Alg.
Deep or Deeper 11
Tianran Zhang | [email protected] | NCEO King's College London
x2 x3
Params/Accuracy
Simple DLResnet20
(n=3)
Params 55,565 858,389
Fire 97% 97%
Land 95% 95%
Cloud 88% 96%
Water 92% 94%
False Fire 67% 84%
12
Tianran Zhang | [email protected] | NCEO King's College London
Potential Application 4:
Smoke Plume IdentificationSEVIRI Active Fire Detections
MODIS AF Detections
MODIS AF Detections Remapped
MODIS 10 km AOD Product
CAMS Global Fire Assimilation System 13
Tianran Zhang | [email protected] | NCEO King's College London