The mission of MATIS Laboratory (Methods of Analysis and Image Processing for Stereo-restitution) is to conduct mathematical and computer research applied to photogrammetry, computer vision and remote sensing for multi-imaging sensors (optical, LiDAR, radar, etc.) and multi-sources.
MATIS’s main missions are to improve the production, analysis and enhancement of IGN repositories through terrestrial, airborne and / or satellite imagery.
Its members have developed skills in the entire imaging chain for remote sensing: radiometry, photogrammetry, physical image analysis and synthesis, 2D and 3D computer vision, three-dimensional reconstruction of surfaces and complex objects, wide indexing scale, pattern recognition, classification, data qualification, visualization and interaction. These skills are deployed in a metrological context specific to photogrammetry, at several levels of the chain. For several years, the variety, complexity and volume of data produced and manipulated have also positioned the laboratory in the field of Big Data, for which it has gained expertise in scaling up his algorithmic methods and in managing large volumes of heterogeneous data.
The MATIS is also involved in the open source community by developing and offering open access libraries and data sets. It also conducts a number of ongoing support activities for IGN production and development services (plot vectorization, 3D model reconstruction of cities, vegetation segmentation and classification, Steréréopolis terrestrial digitization visualization and interaction iTowns, etc.).
- MATIS’s website (fr)
- Real-time Car Detection from Aerial Images in a WebGL World iTowns
Updated the 16 November 2018△