Efficient data-driven gap filling of satellite image time series using deep neural networks with partial convolutions

08/18/2022
by   Marius Appel, et al.
16

The abundance of gaps in satellite image time series often complicates the application of deep learning models such as convolutional neural networks for spatiotemporal modeling. Based on previous work in computer vision on image inpainting, this paper shows how three-dimensional spatiotemporal partial convolutions can be used as layers in neural networks to fill gaps in satellite image time series. To evaluate the approach, we apply a U-Net-like model on incomplete image time series of quasi-global carbon monoxide observations from the Sentinel-5P satellite. Prediction errors were comparable to two considered statistical approaches while computation times for predictions were up to three orders of magnitude faster, making the approach applicable to process large amounts of satellite data. Partial convolutions can be added as layers to other types of neural networks, making it relatively easy to integrate with existing deep learning models. However, the approach does not quantify prediction errors and further research is needed to understand and improve model transferability. The implementation of spatiotemporal partial convolutions and the U-Net-like model is available as open-source software.

READ FULL TEXT

page 6

page 8

page 12

page 13

page 15

research
01/29/2019

Time-Space tradeoff in deep learning models for crop classification on satellite multi-spectral image time series

In this article, we investigate several structured deep learning models ...
research
03/01/2021

Automated data-driven approach for gap filling in the time series using evolutionary learning

Time series analysis is widely used in various fields of science and ind...
research
02/26/2021

Beyond Convolutions: A Novel Deep Learning Approach for Raw Seismic Data Ingestion

Traditional seismic processing workflows (SPW) are expensive, requiring ...
research
07/12/2023

Filling time-series gaps using image techniques: Multidimensional context autoencoder approach for building energy data imputation

Building energy prediction and management has become increasingly import...
research
12/04/2019

Detecting Hardly Visible Roads in Low-Resolution Satellite Time Series Data

Massive amounts of satellite data have been gathered over time, holding ...
research
04/17/2021

Deep Learning in Beyond 5G Networks with Image-based Time-Series Representation

Towards the network innovation, the Beyond Five-Generation (B5G) network...
research
11/28/2021

Learning Wildfire Model from Incomplete State Observations

As wildfires are expected to become more frequent and severe, improved p...

Please sign up or login with your details

Forgot password? Click here to reset