ISPRS Annals of the Photogrammetry, Remote Sensing and Spatial Information Sciences
Download
Publications Copernicus
Download
Citation
Articles | Volume V-2-2022
ISPRS Ann. Photogramm. Remote Sens. Spatial Inf. Sci., V-2-2022, 307–315, 2022
https://doi.org/10.5194/isprs-annals-V-2-2022-307-2022
ISPRS Ann. Photogramm. Remote Sens. Spatial Inf. Sci., V-2-2022, 307–315, 2022
https://doi.org/10.5194/isprs-annals-V-2-2022-307-2022
 
17 May 2022
17 May 2022

DEEP LEARNING FOR THE DETECTION OF EARLY SIGNS FOR FOREST DAMAGE BASED ON SATELLITE IMAGERY

D. Wittich1, F. Rottensteiner1, M. Voelsen1, C. Heipke1, and S. Müller2 D. Wittich et al.
  • 1Institute of Photogrammetry and GeoInformation, Leibniz Universiät Hannover, Germany
  • 2EFTAS Fernerkundung Technologietransfer GmbH, Germany

Keywords: Deep Learning, Regression, Forest Monitoring, Label Imbalance, Sentinel-2

Abstract. We present an approach for detecting early signs for upcoming forest damages by training a Convolutional Neural Network (CNN) for the pixel-wise prediction of the remaining life-time (RLT) of trees in forests based on Sentinel-2 imagery. We focus on a scenario in which reference data are only available for a related task, namely for a bi-temporal pixel-wise classification of forest degradation. This reference is used to train a CNN for the pixel-wise prediction of forest degradation. In this context, we propose a new sub-sampling-based approach for compensating the effects of a heavy class imbalance in the training data. Using the resulting classification model, we predict semi-labels for images of a Sentinel-2 time series, from which training data for a CNN designed to regress the RLT can be derived after some label cleansing. However, due to data gaps in the time series, e.g. caused by clouds, only intervals can be derived for the target variable to be regressed, and for some training pixels one of the interval limits may even be unknown. Consequently, we propose a new loss function for training a CNN for regressing the RLT that only requires the known interval limits. The method is evaluated on a data set in Germany, covering a time-span of 5 years. We show that the proposed sub-sampling strategy for dealing with strong label imbalance when training the classifier significantly reduces the training time compared to other approaches. We further show that our model predicts the RLT with a maximum error of two months for 80% of the forest pixels that die within one year from the acquisition date of the Sentinel-2 image.