Home
Welcome to "Time Series Analysis in Remote Sensing for Understanding Human-Environment Interactions".
The main objective is to learn new tools and methods for time series analysis of remote sensing data and to understand their fit-for-purpose application in Geography and environmental sciences. The focus on this year’s course will be on “3D/4D Geographic Point Cloud Time Series Analysis”.
The course consists of two selected modules of a larger e-learning course called E-TRAINEE that is jointly developed by the Universities of Prague, Warsaw, Innsbruck and Heidelberg in the framework of Erasmus+ and 4EU+ (see https://web.natur.cuni.cz/gis/etrainee/):
- Module 1: This basic module will provide a background on principles of remote sensing time series, time series analysis, time series-based classification, trajectory-based analysis and spatio-temporal data fusion. Furthermore, validation and accuracy assessment principles will be explained.
- Module 3: This core module will deal with 3D/4D geographic point cloud time series analysis. Surface dynamics within a local landscape occur on a large range of spatiotemporal scales. The analysis of surface activities and structural dynamics in 4D point cloud data has therefore become an integral part of Earth observation. These data contain detailed 3D information of the topography with time as additional dimension. This module will also feature Heidelberg’s open source Python library py4dgeo for point cloud change analysis (https://github.com/3dgeo-heidelberg/py4dgeo).
Contact
Prof. Dr. Bernhard Höfle (https://www.geog.uni-heidelberg.de/gis/hoefle_en.html)
Dates and Topics
Location & time
- Hybrid meetings: INF 348 / R13 or heiCONF (link is provided by lecturer)
- 14:15h - end of Q&A and discussions
The dates below represent the due date to finish the topics and be ready for the feedback meeting.
| # | Date | Topics |
|---|---|---|
| 00 | 23.04.2024 | Introduction & Organization |
| 01 | 30.04.2024 | Module 1: Principles of remote sensing time series |
| 02 | 07.05.2024 | Module 1: Large time series datasets in remote sensing |
| 03 | 14.05.2024 | Module 1: Time series analysis based on classification |
| 04 | 21.05.2024 | Module 1: Trajectory-based analysis |
| - | 28.05.2024 | - |
| 05 | 04.06.2024 | Module 1: Spatio-temporal data fusion |
| 06 | 11.06.2024 | Module 1: Reference data, validation and accuracy assessment |
| 07 | 18.06.2024 | Module 3: Principles of 3D/4D geographic point clouds |
| 08 | 25.06.2024 | Module 3: Programming for point cloud analysis with Python |
| 09 | 02.07.2024 | Module 3: Principles and basic algorithms of 3D change detection and analysis |
| 10 | 09.07.2024 | Module 3: Time series analysis of 3D point clouds |
| 11 | 16.07.2024 | Module 3: Machine learning-based 3D/4D point cloud analysis |
| 12 | 23.07.2024 | Module 3: Case studies |
| - | 23.07.2024 | Deadline for the selection of research topic |
| - | 30.09.2024 | Deadline for scientific reports (submission as PDF via email to lecturer) |
How To Use the Course
- The course contents are provided via a website (which you are probably on right now).
- This website is automatically generated from the corresponding Github repository: https://github.com/3dgeo-heidelberg/etrainee_heidelberg_2024. The repository is linked at the top right of the website on all pages.
- Some parts of the course are conducted in Jupyter Notebooks. These are static pages on the website, but it is recommended that you follow them interactively. Use the download button at the top of the respective pages (e.g., here). This will open the raw file in your browser, just right click and use "Save page as / Seite speichern unter..." to store the ipynb file.
- All data required for the course is contained in the central data repository. Each theme contains information which of the data (directories) is relevant for the current task.
- You can report issues via the Issue Tracker on the Github repository. Depending on the type of issue, they will be solved throughout the course or considered in future development.
Set up your Python conda environment
Installation instructions and links to the respective conda environments are given in the software section of this course.
Data Repository
All data used in the course can be accessed via Zenodo and downloaded before starting the respective module: https://zenodo.org/records/10003575