Tutorials

Tutorials at O3DM and LC3D, 18 November 2026

Prior to the workshops, we offer four different tutorials, 2 full-day and 2 half-day, see below.

The two half-day tutorials will take place consecutively. 

Fees:

One registration per full-day tutorial: 200€

One registration per half-day tutorial: 125€

A tutorial will only take place when at least 10 participants are registered. The maximum number of participants per tutorial is 20.

Register here (you will be directed to another site).

Tutorial 1: Hybrid and Precise Camera Pose Estimation in MicMac.v2 (MMVII)

Duration: full-day (8 hours)

Description:
The purpose of this tutorial is to introduce participants to the new version of the open-source MicMac photogrammetric processing chain. This second version (MMVII), currently in development since 2020, offers a metrological refocusing of the processing chain. After a presentation of the processing chain and the philosophy behind its use, participants will proceed to process calibration data (intrinsic camera calibration, rigid block calibration of a multi-camera system) using coded or uncoded photogrammetric targets. Participants will also be able to perform processing using different types of data: for example, hybrid processing combining both measurements from images and conventional measured data from topometric instruments (angles and distances).

Expected number of participants: 10-20

Prerequisites:

• Ability to use a command line interface (CLI)

• Shell for scripting or Python programming.

Hardware & Software:

• Laptop: Participants must bring their own laptop (Windows or Linux).

  • For Linux users please follow instructions here : https://github.com/micmacIGN/micmac/blob/master/MMVII/README.md#linux-ubuntu-distribution

  • Windows users can download pre-compiled binaries here : https://github.com/micmacIGN/micmac/releases/tag/Windows_MMVII_build (we recommand the installation of Git-Bash as a terminal: https://git-scm.com/install/windows).

• Software: Please have a 3D viewer just in case (for example, CloudCompare https://www.cloudcompare.org/ )

• Data Download: Participants will be asked to download the tutorial datasets in advance to ensure a smooth workflow during the session. The link to download the datasets will be shared later.

Tutorial lead:

Mehdi Daakir and Jean-Michael Muller.

Tutorial 2: Hand-Held Mobile Mapping System for Large Scale Surveys: low cost hardware and open source software

Duration: full-day (8 hours)

Description:

This tutorial is introducing an open source (MapsHD/HDMapping), open hardware (https://github.com/JanuszBedkowski/mandeye_controller) projects for building 3D LiDAR mobile mapping system and data processing. It is for large-scale 3D mapping with hand-held/wearable/mountable measurement device. We provide an end-to-end mobile mapping framework that does not require any installation on Window 11, including:

• HDMapping_LI0: our implementation of LiDAR Inertial Odometry that outperforms the State of the Art (timestamp of claim: February 2026).

• HDMapping_Pose_GRAPH_SLAM to create city-level maps.

• HDMapping_Georeferencing (GNSS-RTK, Control Points, Ground Control Points, TLS, ALS).

We provide software tools that integrates following LiDAR odometry algorithms

• https://github.com/MapsHD/benchmark-Super-LIO-to-HDMapping (2026, RA-L, [BIB], [movie])

• https://github.com/MapsHD/benchmark-lidar_odometry_ros_wrapper-to-HDMapping (2025, RA-L, [BIB], [movie])

• https://github.com/MapsHD/benchmark-SuperOdometry-to-HDMapping (2021, IROS, 2025, ICRA, [BIB], [movie])

• https://github.com/MapsHD/benchmark-mola_lidar_odometry-to-HDMapping (2025, IJRR, [BIB], [movie])

• https://github.com/MapsHD/benchmark-RESPLE-to-HDMapping (2025, RA-L, [BIB], [movie])

• https://github.com/MapsHD/benchmark-GenZ-ICP-to-HDMapping (2025, RA-L, [BIB], [movie])

• https://github.com/MapsHD/benchmark-I2EKF-LO-to-HDMapping (2024, arXiv, [BIB], [movie])

• https://github.com/MapsHD/benchmark-LIO-EKF-to-HDMapping (2024, ICRA, [BIB], [movie])

• https://github.com/MapsHD/benchmark-iG-LIO-to-HDMapping (2024, RA-L, [BIB], [movie])

• https://github.com/MapsHD/benchmark-GLIM-to-HDMapping (2024, arXiv, [BIB], [movie])

• https://github.com/MapsHD/benchmark-Point-LIO-to-HDMapping (2024, JAIS, [BIB], [movie])

• https://github.com/MapsHD/benchmark-DLIO-to-HDMapping (2023, ICRA, [BIB], [movie])

• https://github.com/MapsHD/benchmark-SLICT-to-HDMapping (2023, RA-L, [BIB], [movie])

• https://github.com/MapsHD/benchmark-KISS-ICP-to-HDMapping (2023, RA-L, [BIB], [movie])

• https://github.com/MapsHD/benchmark-Faster-LIO-to-HDMapping (2022, RA-L, [BIB], [movie])

• https://github.com/MapsHD/benchmark-VoxelMap-to-HDMapping (2022, arXiv, [BIB], [movie])

• https://github.com/MapsHD/benchmark-DLO-to-HDMapping (2022, RA-L, [BIB], [movie])

• https://github.com/MapsHD/benchmark-CT-ICP-to-HDMapping (2021, arXiv, [BIB], [movie])

• https://github.com/MapsHD/benchmark-FAST-LIO-to-HDMapping (2020, arXiv, [BIB], [movie])

• https://github.com/MapsHD/benchmark-LOAM-Livox-to-HDMapping (2019, arXiv, [BIB], [movie])

• https://github.com/MapsHD/benchmark-LeGO-LOAM-to-HDMapping (2018, IROS, [BIB], [movie])

It is possible to perfrom benchmark of more than 20 recently published LIO algorithms. We support LIVOX AVIA, HAP, MID360, Ouster OS0, OS1, OS2, OSDome, SICK multiScan100, HESAI JT16, HESAI JT128, HESAI XT, Robosense AIRY. HESAI XT requires external IMU. We are going to support more LiDARs ASAP.

During tutorial we will

·         Elaborate recent advances in LiDAR mobile mapping (hardware, software).

·         Collect mobile mapping data with MANDEYE systems.

·         Process mobile mapping data to obtain precise and accurate trajectory and 3D point cloud.

·         Georeference data to global coordinate system.

·         Demonstrate some of alternative LIO algorithms.

·         Guidance to perform a benchmark for different LIO algorithms with different LiDARs.

Expected number of participants: 10-20

Prerequisites:

Hardware & Software:

The project HDMapping is supported for Windows, Linux, MacOS. We recommend the following machines:

·         Laptop with Windows 11, 32GB RAM (can be less e.g. 16GB minimum).

·         Laptop with Ubuntu 24.04, 32GB RAM (can be less e.g. 16GB minimum).

·         Please check if DEMO is working on Your Windows 11 machine HDMapping/quick_start_demo/quick_start_demo.7z at main · MapsHD/HDMapping.

·         Download and check recent release Release v0.98 · MapsHD/HDMapping.

·         If You are using Ubuntu/MacOS please follow the building instructions at MapsHD/HDMapping.

Tutorial lead:

Janusz Będkowski, PhD, DSc

Tutorial 3: Open-source scientific software py4dgeo for change analysis in 3D/4D point clouds

Duration: half-day (3.5 hours)

Description:

This tutorial introduces py4dgeo, an open-source Python library for analyzing geometric change and surface dynamics in 3D and 4D point cloud data. Designed to support scientific workflows, py4dgeo provides a reproducible and scalable tool for quantifying surface change in multitemporal point clouds and 3D time series, for a broad range of applications of topographic monitoring. Participants will gain a solid understanding of the key concepts and challenges in 3D/4D change analysis. This includes the full pipeline of multi-temporal point cloud alignment, 3D change quantification, and time series-based quantification methods. Participants will learn how py4dgeo implements state-of-the-art algorithms to address these challanges. The tutorial emphasizes the modular framework with easily accessible and configurable methods, providing users a powerful toolset for transparent, fully automated scientific analysis. Hands-on exercises will guide participants through practical use of the library, including loading point cloud data, applying 3D change detection algorithms (e.g., M3C2), using a hierarchical approach for 3D change analysis, performing time series-based analysis (e.g., 4D objects-by-change), and visualizing results. Example workflows will show how py4dgeo integrates into common Python-based environments and complements other open-source tools such as CloudCompare. By the end of the tutorial, participants will be able to understand core methods for 3D/4D change analysis, reproduce and adapt workflows for research or applied monitoring tasks, and apply py4dgeo to their own datasets. This tutorial is intended for researchers, students, and practitioners working with time-dependent 3D data in need of a flexible, scalable, and open software framework for surface change analysis.

Expected number of participants: 10-20

Prerequisites:

• A basic understanding of topographic point clouds is required.

• Familiarity with basic Python scripting is required to follow the practical exercises.

Hardware & Software:

• Laptop: Participants must bring their own laptop (Windows or Linux).

• Python Environment: Python >=3.9, conda with installation environment recommended, specific setup instructions will be provided in advance.

• Software: Please have CloudCompare (https://www.cloudcompare.org/) installed.

• Data Download: Participants will be asked to download the tutorial datasets (max. 1 GB) in advance to ensure a smooth workflow during the session.

Tutorial lead:

The tutorial is offered by the TUM Professorship of Remote Sensing Application (TUM-RSA) under the lead of Xiaoyu Huang (https://www.linkedin.com/in/xiaoyu-huang-19a581244/), a researcher specializing in 3D/4D change detection, point cloud processing, and geospatial data analysis. More information about the hosting group can be found at: https://www.asg.ed.tum.de/en/rsa/home/

Tutorial 4: GEOMAPI for construction site monitoring with close-range sensing

Duration: half-day (3.5 hours)

Description:

This workshop, organized by Maarten Bassier in collaboration with Frédéric Bosché, focuses on automation in construction site monitoring through the integrated use of point clouds, imagery, and BIM data. The session addresses one of the key challenges in digital construction workflows: how to efficiently and reliably assess construction progress and geometric quality of structural elements using multi-sensor data.

Participants will work with the GEOMAPI API developed by the Geomatics Group (KU Leuven), exploring how to load and manage BIM objects, geolocated imagery, and point cloud data within a unified framework. By jointly processing these heterogeneous data sources, we will compute indicators related to construction progress and dimensional compliance of built elements.

A central component of the workshop is the graph-based node system underpinning GEOMAPI. This system enables structured storage and propagation of metadata across measurement epochs. Participants will learn how uncertainty information can be transferred between successive surveys, and how previous detections can be leveraged to constrain the search space within large-scale construction site point clouds. This approach significantly improves computational efficiency and robustness when dealing with vast and complex datasets.

In addition, we will examine common close-range sensing deficiencies encountered on construction sites, such as occlusions, incomplete coverage, and varying point densities. Through practical examples, participants will investigate how these limitations influence progress estimation and tolerance verification of structural components, and how uncertainty-aware processing strategies can mitigate their impact.

The workshop combines conceptual explanations with hands-on exercises in Python, guiding participants through loading datasets, querying BIM elements, linking them to sensor observations, and computing geometric and quality metrics. By the end of the session, participants will understand how to set up automated, reproducible workflows for construction monitoring and tolerance control using open and research-driven tools.

Expected number of participants: 10-20

Prerequisites:

• A basic understanding of topographic point clouds is required.

• Familiarity with basic Python scripting is required to follow the practical exercises.

Hardware & Software:

• Laptop: Participants must bring their own laptop (Windows or Linux).

• Python Environment: Python >=3.9, conda with installation environment recommended, specific setup instructions will be provided in advance.

• Software: Please have CloudCompare (https://www.cloudcompare.org/) installed.

• Data Download: Participants will be asked to download the tutorial datasets (max. 1 GB) in advance to ensure a smooth workflow during the session.

Tutorial lead:

The tutorial is offered by Maarten Bassier and Frédéric Bosché in collaboration with the Geomatics Group (KU Leuven). https://ku-leuven-geomatics.github.io/geomapi/