The image processing team is involved in developing open-source libraries to provide reusable piece of efficient image processing software. More specifically, we recently addressed the following challenges:
- Providing a mathematical morphology-related Modern C++ library that conciliates ease-of-use, performance and genericity (see pylene)
- Developing high-performance mathematical morphology algorithms (mostly related to hierarchical image representations) that exploit the massively parallel architectures (e.g. GPUs) to gain efficiency and achieve state-of-the-art performance in this field.
Generic image processing with Pylene: an algorithm is written once to process many kinds of images (regular 2D images, graphs and meshes).
Main contact: Edwin Carlinet
