Research Axes

In collaboration with SAFRAN Reosc and the Laboratoire d'Astrophysique de Marseille (CNRS-LAM), we are studying the impact of free-form optics and curved sensors in the design of a wide-field imaging system for situational awareness applications in the visible and near-infrared spectrum. In collaboration with Saint Gobain Recherche (SGR), we are also studying the design of free-form optics for automotive and aerospace windshields. In collaboration with CEA-LETI and Ch. Sauvan (Nanophotonics team), we are studying the design of freeform components 'dressed' with metasurfaces to improve the coupling of light between sources and integrated photonic circuits. In collaboration with the French Freeform Optics Research Association (FO-RS), we are also studying the metrology of various freeform surfaces (mirrors, ophthalmic optics, LED-based lighting systems, etc.). Finally, we are studying new ways of manufacturing optical components, whether free-form or not, at the LCF (lenses, mirrors) and also in collaboration with the CEA (Saclay) and SARAN Reosc (SiC mirrors).

The bidirectional reflection distribution function (BRDF) is known to play an important role in the appearance of certain surfaces, as it depends on their roughness and irregularities. The BRDF also depends on the angular resolution of the detection device, which can be high in certain industrial applications (cosmetics, fashion, watchmaking, etc.). In collaboration with the CNAM (Paris, G. Obein's team), the Institut P' (Poitiers, L. Simonot's team) and the Laboratoire Hubert Curien (Saint Etienne, M. Hébert's team), we are studying how increasing the angular resolution affects the measurement of the BRDF. At the same time, we are characterising the appearance of surfaces by directly measuring the roughness and irregularities of material surfaces. This latter activity has been and is supported by contracts with Chanel Parfums Beauté and SGD Pharma.

Artificial intelligence (AI) is known to accelerate decision-making. It is therefore a powerful in many applications, such as the automated control of driving offences. However, it is based on training neural networks with images, the quality of which is essential to limit detection errors. In collaboration with FARECO (FAYAT group) and Matthieu Boffety (Imagerie & Information team at LCF), we investigate various means to improve optronic systems dedicated to such automated control, accounting for improved illumination, imaging optics, and detection system.