Bio Image Analysis.ORG

Description : In the field of material science, particles of nanometric dimensions (nanoparticles) have fascinating properties stemming from their size, shape, among other things, which give them wide applications in nanotechnology. Characterization of the size and shape of these nanoparticles at the nanoscopic scale is based on electron microscopy images. The physical properties (e.g. magnetic or optical) of these nanomaterials are characterized using various physico-chemical techniques that provide data (spectra, curves, etc.). In this way, each nanomaterial is characterized by its size, shape and properties. The aim of this project is to develop image analysis methods adapted to this problem.

Pré-requis : Image analysis and machine learning

Travail demandé : 1. Literature review. 2. Development of an image analysis method for nanoparticles. The first step is to segment and separate the particles. Depending on the amount of data available, structural methods (e.g. mathematical morphology) or learning methods can be developed. Next, geometric characteristics such as size, orientation and anisotropy can be calculated from the segmentation results. Finally, depending on the time available, predictive methods can be developed, such as learning from multiple matched modalities (images, spectra) to infer one modality from the other once learning is complete.