Alice Fontbonne

Alice Fontbonne

I am a researcher at ONERA since 2023

Research projects

My research is in the field of optical design and joint optical/digital processing design. It can be separate into two global research projects.

Joint optical/digital processing design

Today, most imaging systems consist of an optical system and digital processing. For the imaging system to be optimal (while respecting weight and power constraints), these two elements must be optimized at the same time. However, for these two parts of the imaging system to be jointly optimized, it is necessary to develop and use dedicated software, such as a differentiable ray tracer. In this way, the imaging characteristics (point spread function) taken into account in the digital processing are linked to the derivatives of the optical parameters (radius of curvature, thickness…). This makes possible to consider different digital processing algorithms, ranging from linear deconvolution to learning approaches.

Co-design strategies exploration

Since 2025, I am the project coordinator of ANR JCJC EDDRIS project. For the next few years, we propose to use the ESCL-licensed FORMIDABLE differentiable ray tracer, and to develop an environment specifically dedicated to joint optical/processing design. In the course of the project, we will be looking at imaging systems of increasing complexity (using phase masks in the pupil of the optical device to modify the wavefront, then modifying a larger number of surfaces, possibly freeform). We will propose specific design strategies for the different types of processing envisaged, in line with the size, weight and power consumption constraints of the imaging system. A demonstrator dedicated to car driving will enable us to validate experimentally one of the strategies we have put in place. Find more information about this project on EDDRIS’s website!

PhD Students supervision:

Deep Co-design of a Privacy Preserving Intelligent Camera

Since 2025, I am part of ANR DOOPLER project, leads by Pauline Trouvé-Peloux. Today, we are surrounded by surveillance systems that generate billions of images per day processed by computer vision techniques. This situation raises legal issues related to privacy. If privacy is sometimes protected by post-processing the images, for example by blurring them, this protection can be attacked by gaining access to the camera’s raw output. The challenge is therefore to provide privacy directly at the camera level by deliberately degrading the quality of the images produced by the lens. We propose the deep co-design of a privacy preserving camera using a differentiable ray tracing software developed at ONERA.

PhD Students supervision:

Freeform optics for compactness & optimization

When joining ONERA, I’ve been part of Clément Freslier’s supervision team. His work focuses on compact imaging optical systems embedded into small satellites. Achieving good optical performance with exemplary compactness requires the use of freeform mirrors. We focused on the NURBS (Non Uniform Rational B-splines) surface representation, implemented in FORMIDABLE, to describe and optimize such systems. His thesis highlighted the need to focus on the process of optimising such optical systems: are the algorithms typically used in optical design still suitable for such problems?

PhD Students supervision: