Digital Terrain Modelling

Digital terrain models (DTMs) are computer representations of the elevation of a given area, and they play an important role in understanding and analysing our built environment. They are the necessary input for several applications (eg flood modelling, visibility, effects of climate change on the north poles, etc.), and they are also relevant for studying the seabed.

The course provides an overview of the fundamentals of digital terrain modelling:

- different representations of terrains: TINs, rasters, point clouds, contour lines

- reconstruction of terrains from different sources (lidar, InSAR, photogrammetry, multibeam echosounders)

- spatial interpolation methods

- conversion between different representations

- processing of terrains and point clouds: outlier detection, filtering, segmentation, and identification and classification of objects

- techniques to handle and process massive datasets

- applications, eg runoff modelling, watersheds computations, visibility

The course has both a theoretical part and a practical part where students reconstruct, manipulate, process, and extract information from terrains.

All the labs are programming tasks (to be done with Python and/or C++), based solely on open-source libraries and software.

- two written exams (1 mid-term (at week 2.5) + 1 final exam)

- assignments (programming in Python and/or C++)

- there are 2 retakes (during Q3): one for both written exams + a retake assignment for the assignments

The weights for the exams and the assignments are around 50%/50%.

Assignments submitted after the deadline: the grade will be deducted with 10% per day late, and not accepted after 3 days late.

A mark of 5.00 in the exams part and a total course mark of 5.75, are both necessary to pass the course.