Country-wide retrieval of forest structure from optical and SAR satellite imagery with deep ensembles

Becker, Alexander; Russo, Stefania; Puliti, Stefano; Lang, Nico; Schindler, Konrad; Wegner, Jan Dirk

doi:10.1016/j.isprsjprs.2022.11.011

Cited by 22 publications

(9 citation statements)

References 73 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Regarding coupling of ALS data with satellite EO data in our pretraining in Lapland, our results are in line with other similar studies [41]. Use of recurrent and fully convolutional neural networks with fully segmented labels and Sentinel-1 time series or combined SAR and optical data provided accuracies on the order of 17-30% rRMSE that are similar to results in our work over pretraining site in Lapland [15], [16], [20], [42]. Inversion of TanDEM-X images acquired over Estonian hemiboreal and Canadian boreal forests provided accuracies with RMSE in range of 3-4 m and correlation coefficients R 2 larger than 0.5 [40], [43]- [46].…”

Section: B Comparison With Prior Studiessupporting

confidence: 91%

Deep Learning Models in Forest Mapping Using Multitemporal SAR and Optical Satellite Data

et al. 2022

IGARSS 2022 - 2022 IEEE International Geoscience and Remote Sensing Symposium

View full text Add to dashboard Cite

Deep learning (DL) models are gaining popularity in forest variable prediction using Earth Observation images. However, in practical forest inventories, reference datasets are often represented by plot-or stand-level measurements, while high-quality representative wall-to-wall reference data for end-toend training of DL models are rarely available. Transfer learning facilitates expansion of the use of deep learning models into areas with sub-optimal training data by allowing pretraining of the model in areas where high-quality teaching data are available. In this study, we perform a "model transfer" (or domain adaptation) of a pretrained DL model into a target area using plot-level measurements and compare performance versus other machine learning models. We use an earlier developed UNet based model (SeUNet) to demonstrate the approach on two distinct taiga sites with varying forest structure and composition. Multisource Earth Observation (EO) data are represented by a combination of Copernicus Sentinel-1 C-band SAR and Sentinel-2 multispectral images, JAXA ALOS-2 PALSAR-2 SAR mosaic and TanDEM-X bistatic interferometric radar data. The training study site is located in Finnish Lapland, while the target site is located in Southern Finland. By leveraging transfer learning, the prediction of SeUNet achieved root mean squared error (RMSE) of 2.70 m and R 2 of 0.882, considerably more accurate than traditional benchmark methods. We expect such forest-specific DL model transfer can be suitable also for other forest variables and other EO data sources that are sensitive to forest structure.

show abstract

Section: B Comparison With Prior Studiessupporting

confidence: 91%

Deep Learning Models in Forest Mapping Using Multitemporal SAR and Optical Satellite Data

et al. 2022

IGARSS 2022 - 2022 IEEE International Geoscience and Remote Sensing Symposium

View full text Add to dashboard Cite

show abstract

“…It does that astonishingly well, with a MAE around 1 5000 of the intensity range-less than the radiometric sensitivity of Sentinel-2. 7 The estimates of U-TILISE are also qualitatively superior to those of the baselines, especially in the presence of significant spectral changes in time (cf. Fig.…”

Section: B Comparison To Baselinesmentioning

confidence: 99%

U-TILISE: A Sequence-to-Sequence Model for Cloud Removal in Optical Satellite Time Series

Stucker,

Garnot,

Schindler

2023

IEEE Trans. Geosci. Remote Sensing

Self Cite

View full text Add to dashboard Cite

Satellite image time series in the optical and infrared spectrum suffer from frequent data gaps due to cloud cover, cloud shadows, and temporary sensor outages. It has been a long-standing problem of remote sensing research how to best reconstruct the missing pixel values and obtain complete, cloud-free image sequences. We approach that problem from the perspective of representation learning and develop U-TILISE, an efficient neural model that is able to implicitly capture spatio-temporal patterns of the spectral intensities, and that can therefore be trained to map a cloud-masked input sequence to a cloud-free output sequence. The model consists of a convolutional spatial encoder that maps each individual frame of the input sequence to a latent encoding; an attention-based temporal encoder that captures dependencies between those per-frame encodings and lets them exchange information along the time dimension; and a convolutional spatial decoder that decodes the latent embeddings back into multi-spectral images. We experimentally evaluate the proposed model on EarthNet2021, a dataset of Sentinel-2 time series acquired all over Europe, and demonstrate its superior ability to reconstruct the missing pixels. Compared to a standard interpolation baseline, it increases the PSNR by 1.8 dB at previously seen locations and by 1.3 dB at unseen locations.

show abstract

“…Inventories provide essential information on forest biomass stocks used for climate treaties and carbon accounting but are time-consuming, labor-intensive, and limited to plot scale, and the methods, and degree to which monitoring of trees outside forests is conducted, vary substantially across countries ( 9 , 10 ). Comprehensive information on forests, such as forest cover ( 11 ), structure ( 12 ), resources ( 13 , 14 ), phenology ( 15 , 16 ), disturbances ( 17 ), and diversity ( 14 , 18 , 19 ) at national scale, is commonly derived from remote sensing data, often combined with inventory measurements ( 20 ). Satellite-based monitoring of forests based on readily available satellite data with a spatial resolution down to 10 m enables low-cost ( 21 ) and wall-to-wall assessments that can be rapidly repeated at a high temporal frequency and a large scale ( 11 ).…”

Section: Introductionmentioning

confidence: 99%

Deep learning enables image-based tree counting, crown segmentation, and height prediction at national scale

Brandt

Fensholt

et al. 2023

PNAS Nexus

View full text Add to dashboard Cite

Sustainable tree resource management is the key to mitigating climate warming, fostering a green economy, and protecting valuable habitats. Detailed knowledge about tree resources is a prerequisite for such management but is conventionally based on plot-scale data, which often neglects trees outside forests. Here, we present a deep learning-based framework that provides location, crown area and height for individual overstory trees from aerial images at country scale. We apply the framework on data covering Denmark and show that large trees (stem diameter > 10 cm) can be identified with a low bias (12.5%), and that trees outside forests contribute to 30% of the total tree cover, which is typically unrecognized in national inventories. The bias is high (46.6%) when our results are evaluated against all trees taller than 1.3 m, which involve undetectable small or understory trees. Furthermore, we demonstrate that only marginal effort is needed to transfer our framework to data from Finland, despite markedly dissimilar data sources. Our work lays the foundation for digitalized national databases, where large trees are spatially traceable and manageable.

show abstract

Country-wide retrieval of forest structure from optical and SAR satellite imagery with deep ensembles

Cited by 22 publications

References 73 publications

Deep Learning Models in Forest Mapping Using Multitemporal SAR and Optical Satellite Data

Deep Learning Models in Forest Mapping Using Multitemporal SAR and Optical Satellite Data

U-TILISE: A Sequence-to-Sequence Model for Cloud Removal in Optical Satellite Time Series

Deep learning enables image-based tree counting, crown segmentation, and height prediction at national scale

Contact Info

Product

Resources

About