Research themes

Great deal of problems revolve around three main themes:

Biomedical imaging

Biology and medical image analysis, radiological image mining

See related projects See publications
Document image

Symbol recognition, document analysis and understanding

See related projects See publications
Remote sensing imaging

Aerial image processing, high resolution satellite image interpretation

See related projects See publications

We address the structural models (graph or pixel grid) as well as the statistical models (distributions) by attaching us to integrate external knowledge to the image analysis workflow (complementary textual information and high-level expert knowledge for example). In the context of image mining, recognition of complex shapes, feature indexing and selection can characterize our research activities. The goal is to develop methods from pattern recognition, image analysis and artificial intelligence theory to provide functional and original solutions to different problems related to visual perception.