Day 1, Monday, October 21

Lecture 1: Introducing Laser Scanning, 8:00 - 9.45

This lesson will introduce principles of laser scanning, which is used to capture 3D topographic data of surfaces. This lecture discusses different measurement principles and different systems of data acquisition. Both current and upcoming techniques will be presented. The lesson will cover:

• Laser scanning technology
• Measurement principles: phase shift, time of flight (pulse-base) and photon counting
• Discrete and full waveform laser scanning
• Laser scanning systems: terrestrial laser scanning, mobile laser scanning, backpack and aerial laser scanning (UAV, fixed wings and helicopter)
• Laser scanning intensity and multispectral

Exercise 1: Data acquisition (field work), 10:00 - 11:45

In this session, students will use the terrestrial laser scanner to capture an exterior building and road section or trees

Lecture 2: Laser Scanning Processing 1, 13:00 - 15:15

This section will cover the main steps to process raw point cloud data, and will include:

• Data registration: principles of data registration, iterative closest point (ICP), target-based and features based registration
• Geo-referencing: Control points, GPS , direct georeference, and geoslam
• Data errors: outliers, bias occlusion and noise
• Data filtering: outlier removal (e.g. statistical outlier removal), mixed pixel removal (e.g. intensity or color based)
• Subsampling data: space, octree, random

Exercise 2: Laser Scanning Processing 1, 15:30 - 17:15

List of exercises: Appendix.pptx

This section will demo how to use the software to register and clean data for further processing. Students can use open sources software for this demonstration.

• Demonstrate to use artificial targets and object features to register TLS data (use a commercial software)
• Outlier removal (use open sources likely CloudCompare)
• Subsampling data (use open sources likely CloudCompare)

Day 2, Tuesday, October 22

Lecture 3, Laser Scanning Processing 2, 08:00 - 9:45

This lecture introduces processing methods to extract useful information from raw point clouds. The methods covered include:

• Segmentation: Region growing, Model fitting (Hough Transform and RANSAC) and Machine Learning
• Classification: handcrafted and learned features, and supervised, unsupervised and deep learning
• Feature extraction: Linearity, planarity and scatterity, edge detection, boundary extraction, line, plane, cylinder, sphere

Exercise 3, Laser Scanning Processing 2, 10:00 - 11:45

This classroom exercise class uses open sources software or programming language to compute features of the points cloud for segmentation and practice on segmentation methods.

• Compute features of the points (Use either Python, Matlab or open source as Cloud Compare)
• Segmentation by RANSAC (e.g. Cloud Compare)
• Fitting lines, plane, cylinder (Matlab or Python).

Lecture 4, Laser Scanning Processing 3, 13:30 - 15:15

This lecture will present methods for modelling a surdace, starting from a point cloud, and will include:

• (1) Quadtree and Octree representation; (2) Voxel grid representation; (3) Constructive solid geometry (CSG)
• Geometric modelling: (1) planar surface; (2) curved surface; (3) Extrusions; (4) Sweeping
• Mesh models: 2D and 3D Triangulation mesh

Exercise 4, Laser Scanning Processing 3, 15:30 - 17:15

The class exercise demonstrates how to process point clouds for 3D model generation for civil engineering application. This exercise strongly focusses on creating 3D model for BIM. The participants will create 3D structural model for BIM (Archicad, Revit)

Day 3, Wednesday, October 23

Lecture 5, LIDAR classification & Point cloud applications

• Point cloud classification using e.g. Random Forest, Lecturer: Roderik Lindenbergh ; Slides: vietnamLS2classf.pdf, 08:00 - 08:45
• Change Detection, Lecturer: Roderik Lindenbergh ; Slides: vietnamLS5Bchange.pdf, 09.00 - 09.45
• Contours, excavation monitoring and deformation, Lecturer: Linh Truong-Hong, Slides: PC_VN_Lecture_6b.pdf 10.00 - 10.45
• Surface damage detection, Linh Truong-Hong, 11.00 - 11.45

Lecture 6, UAV applications

This section will present application of UAV for structural monitoring and environmental engineering.

• Drones for surveying and inspection, Linh Truong-Hong, 13.30 - 14.15
• Drones for agriculture, Thu Anh Pham, 14.30 - 15.15

Exercise 6, Point cloud applications \& UAV, 15:30 - 17:15

The class exercise demonstrates how to process point clouds for civil engineering and environmental applications:

• DEM
• Contours (an open source i.e. Cloud Compare)
• Change detection (Matlab, Python or Cloud Compare)

Some Point Cloud data

• TLS data, Delft windmill, `De Roos`
• Rock glacier TLS data, Obergurgl
• Mountain valley, TLS data, Obergurgl
• TLS data of Swedish trees
• AHN airborne LIDAR data
• Flamish Belgium, airborne LIDAR data
• Montreal, airborne LIDAR
• Paris, mobile mapping data
• Annotated Dublin ALS

Laser scanning – for environmental and structural monitoring