Unsupervised LSTM-AE and Harmonic Models for Improving Timeliness of Forest Logging Notifications With Dense Satellite Image Time Series

Keyphrases

• Finland 100%
• Forest Logging 100%
• Change Map 100%
• Image Time Series Analysis 100%
• Harmonic Model 100%
• Long Short-term Memory Autoencoder 100%
• Autoencoder Model 100%
• Dense Satellite Image Time Series 100%
• Satellite Image Time Series 80%
• Logging Operations 60%
• Storm Damage 60%
• Change Classification 60%
• Continuous Change Detection 60%
• Forest Change 60%
• Spatial Accuracy 60%
• Time Series Analysis Method 60%
• Accuracy Assessment 40%
• Bitemporal 40%
• Series Analysis 40%
• High Resolution Satellite Images 40%
• Magnitude of Change 40%
• Overlapping Area 40%
• Landsat Images 40%
• Temporal Change Detection 40%
• Continuous Change 20%
• Euro 20%
• Operating Conditions 20%
• Deforestation 20%
• Power Lines 20%
• Long Delay 20%
• Fully-automatic 20%
• On Demand 20%
• Electronic Health Records 20%
• LiDAR 20%
• Classification Algorithms 20%
• Forest Research 20%
• High-resolution Image 20%
• Short-wave Infrared 20%
• Air Force 20%
• Automatic Analysis 20%
• Severe Damage 20%
• Forest Plot 20%
• Infrared Band 20%
• High Latitude 20%
• Carbon Stock 20%
• Forest Owners 20%
• Railroad 20%
• Time Series Data 20%
• Operational Practices 20%
• AstA 20%
• Damage Mapping 20%
• Financial Loss 20%
• Supervised Classification 20%
• Change Detection Method 20%
• Forest Degradation 20%
• Skipping 20%
• Cloud Cover 20%
• Systematic Bias 20%
• Radiometric Correction 20%
• Image Time Series 20%
• Landsat 20%
• Forest Damage 20%
• Acquisition Frequency 20%
• Concise Representation 20%
• Dense Time Series 20%
• Change Detection Algorithms 20%
• Change Monitoring 20%
• Field Visit 20%
• Temporal Accuracy 20%
• Processing Platform 20%
• Sentinel-2 Satellite 20%
• Annual Time Series 20%
• Biased Randomization 20%
• Autoencoder 20%
• Flyover 20%
• Detection Parameters 20%
• Damage Map 20%
• Landsat Archive 20%
• Storm Events 20%
• Temporal Sampling 20%
• Drive-by 20%
• Temporal Domain 20%
• Time Representation 20%
• Major Storms 20%
• Long-short Term Memory Cell 20%
• Map Updating 20%
• Sun Angle 20%
• Temporal Sparsity 20%
• Missing Observations 20%
• Classification Parameters 20%
• Image Stack 20%
• Private Owner 20%

INIS

• images 100%
• satellites 100%
• forests 100%
• harmonics 100%
• timeliness 100%
• maps 58%
• time-series analysis 41%
• data 35%
• detection 35%
• storms 35%
• finland 29%
• damage 29%
• accuracy 29%
• operation 23%
• classification 23%
• resolution 17%
• assessments 11%
• mapping 11%
• comparative evaluations 5%
• information 5%
• production 5%
• cost 5%
• carbon 5%
• processing 5%
• tolerance 5%
• power 5%
• demand 5%
• monitoring 5%
• losses 5%
• sampling 5%
• air 5%
• stocks 5%
• availability 5%
• algorithms 5%
• stacks 5%
• randomness 5%
• filters 5%
• sun 5%
• cutting 5%
• corrections 5%
• deforestation 5%
• lidar 5%
• cloud cover 5%

Earth and Planetary Sciences

• Satellite Image 100%
• Time Series 100%
• Harmonics 100%
• Change Detection 71%
• Finland 71%
• Clearcutting 57%
• Landsat 57%
• Storm Damage 42%
• Satellite Imagery 28%
• Accuracy Assessment 28%
• State of the Art 14%
• Detection Method 14%
• Preprocessing 14%
• Power Line 14%
• Cloud Cover 14%
• Deforestation 14%
• Supervised Classification 14%
• Sentinel-2 14%
• Continent 14%
• Carbon Sink 14%
• Radiometric Correction 14%
• Optical Radar 14%