PhD position in geomatics science

Machine learning on surface networks for landform classification

Département des Sciences Géomatiques
Centre de Recherche en Données et Intelligence Géospatiales
Université Laval

Context and objective
Landforms are features of the earth’s surface that are easily recognisable by their geometrical shape. Many studies have focused on their automatic recognition from digital terrain models. Current approaches, including approaches based on deep learning, are often based on image segmentation techniques. However, landforms are objects that are difficult to delineate and their definition is based on the perception of prominences (ridges, summits, etc.) by the user.
In earlier work, the team developed an approach extracting the network of ridges and valleys connecting the peaks and valleys of a terrain. Considering that this network contains the prominences of the terrain, the objective of this thesis is to develop an approach classifying landforms directly from the network. Methods based on unsupervised or semi-supervised learning are considered given the difficulty of having previously labeled datasets.

The candidate must have a profile related to spatial data modelling with a master’s degree in geomatics, computer science or a related field. Knowledge of machine learning techniques is an advantage. The student will be required to use development libraries such as Tensor Flow or PyTorch.

Fundings is guaranteed by a NSERC discovery grant for three years. A scholarship of 22 500$/year will be allocated to the student.

Workplace and language
This PhD thesis will take place in Québec City, on the main campus of Université Laval. The student will be registered in the PhD programme in geomatics science and will be a member of the Research Centre in Geospatial Data and Intelligence. Université Laval is a French speaking university however the research work can be conducted in English.

Beginning of the project :
January 2023 or as soon as possible
To apply, please send to Éric Guilbert (
• A cover letter related to the position,
• your curriculum vitae,
• your most recent transcript,
• 2 references

This post is also available in: French