2nd International Workshop
Point Cloud Processing


Dec 4-5, 2019


Stuttgart, Germany

About The Event

2nd International Workshop “Point Cloud Processing"

The 2nd International Workshop on Point Cloud Processing was co-organized by EuroSDR and the Institute for Photogrammetry. There were about 90 participants from 21 nationalities following 20 presentations about point cloud acquisition, processing, classification and interpretation.

Topics | Focus

The quality of area covering 3D point clouds as captured by aerial and mobile mapping platforms still experiences a considerable boost due to the ongoing advancements in LiDAR technology and Multi-View-Stereo-Matching (MVS). In addition to further enhancement of the respective accuracy, density and reliability the semantic segmentation of these point clouds come more and more into focus. Also triggered by the astonishing improvements in the field of pattern recognition and machine learning, automatic interpretation of area covering point clouds including tasks like object detection and classification or object-dependent filtering and smoothing is moving rapidly to a mature state. In view of these developments, the meeting brings together experts from industry, academia and national mapping agencies to present and discuss the processing and evaluation of point clouds focusing on mapping purposes.

The program will provide a mix of invited speakers from industry, academia and governmental organizations as well as presentations selected on an abstract based review process.

Themes of the event include

  • acquisition & visualisation
  • data structures & management,
  • segmentation & classification,
  • object detection & reconstruction,
  • applications & use-cases from practice

Event | Organizers

Norbert Haala

Norbert Haala

University of Stuttgart

Uwe Sörgel

Uwe Sörgel

University of Stuttgart

Michael Cramer

Michael Cramer

University of Stuttgart

Fabio Remondino

Fabio Remondino

FBK Trento, Italy

Event | program

Program (pdf)

Wednesday, December 4th

10:30 – 17:00 Registration
Fee includes coffee breaks, social dinner at Wednesday and lunch at Thursday


11:30 – 13:10 Data collection and processing


Gottfried Mandlburger, TU Wien
Modern LIDAR Sensor Technologies – an Overview
Abstract | Presentation

Boris Jutzi, Karlsruhe Institute of Technology
Self-Calibration for LiDAR-based Mobile Mapping Systems
Abstract | Presentation

Konrad Wenzel, Tobias Hauck, nFrames
Dense Image Matching and LiDAR for large scale surface reconstruction
Abstract | Presentation

Daniel Girardeau-Montaut, CloudCompare
Point processing with CloudCompare
Abstract | Presentation


13:10 – 13:40 Coffee Break


13:40 - 15:00 Deep Learning and 3D Point Clouds

Martin Weinmann, Karlsruhe Institute of Technology
Semantic Segmentation of Dense Point Clouds
Abstract | Presentation

Sören Discher, Hasso Plattner Institute, Berlin, Point Cloud Technology GmbH
Deep Learning for Point Cloud Analysis and Classification
Abstract | Presentation

Loic Landrieu, IGN/LASTIG MATIS, Université Paris Est
Deep Learning for 3D Point Clouds
Abstract | Presentation


15:00 – 15:30 Coffee Break


15:30 - 17:10 Training and Reference Data

Michael Kölle, University of Stuttgart
Crowd-Based Active Learning on 3D Point Clouds
Abstract | Presentation

Sander Oude Elberink, Zhishuang Yang, University of Twente
Using training samples retrieved from a topographic map for the classification of Airborne Laser Scanner data
Abstract | Presentation

Yuxing Xie, Jiaojiao Tian, Xiao Xiang Zhu, Peter Reinartz, German Aerospace Center (DLR) Oberpfaffenhofen
Deep Learning-based Point Cloud Classification on Multiple Airborne LiDAR Datasets
Abstract | Presentation

Florian Gandor, Andre Streilein, Swisstopo
Point cloud processing at a national scale – advantages and challenges
Abstract | Presentation


19:00 – 23:00
Social dinner at Restaurant https://www.schlachthof-stuttgart.de/


Thursday, December 5th

8:00 – 15:00 Registration
Fee includes coffee breaks, social dinner at Wednesday and lunch at Thursday

8:30 – 9:50 Data structures and accuracies

Dominik Laupheimer, University of Stuttgart
Deriving Semantics from Textured Meshes
Abstract | Presentation

Florent Poux, University of Liège
Smart Point Cloud

Abstract | Presentation

Lukas Winiwarter, Bernhard Höfle, University of Heidelberg
Error budget of laser scanners: Applications in geomorphic surface change quantification
Abstract | Presentation


9:50 – 10:20 Coffee Break


10:20 – 12:10 Application I:  Buildings, Trees, Bridges

Liangliang Nan, TU Delft
Accurate, detailed, and automatic reconstruction of real-world trees from point clouds
Abstract | Presentation

Stefan Schmohl, University of Stuttgart
Sparse 3D CNNs for Citywide Semantic Mapping with ALS
Abstract | Presentation

Valentina Schmidt, Martin Kada, TU Berlin
Towards a Deep Learning Framework for 3D Building Reconstruction
Abstract | Presentation

Sara Lorite Martínez, Jesús Moreno Jabato, Jesús María Garrido Sáenz de Tejada, Borja Rodríguez-Cuenca,  Instituto Geográfico Nacional, Madrid
Automatic Classification of Bridges and Continental Water Bodies from 3D Point Clouds

Abstract | Presentation


12:10 – 13:10 Lunch
  included in registration fee 


13:10 – 14:30 Application II: Buildings, Powerlines and Visualisation

Ravi Peters, TU Delft
LoD2 building reconstruction from airborne point clouds
Abstract | Presentation

Nan Li¹, Olaf Kaehler², Norbert Pfeifer¹, ¹TU Wien, ²Research Group Active Vision Technologies, Siemens AG Österreich
LiDAR point classification of power line facilities using the deep neutral network PointNet++

Abstract | Presentation

Michael Weinmann, University of Bonn
Let's be there - Sharing Immersive Live Telepresence Experiences based on Efficient Real-time 3D Reconstruction and Streaming
Abstract | Presentation

Closing of workshop


15:00 – 16:00 Add on: Hands-on demo with CloudCompare


Registration fee includes coffee breaks, social dinner at Wednesday, December 4th and lunch at Thursday, December 5th

Promo | video