Responsable : Adrian BASARAB
MINDS is a research group that conducts research activities in computational imaging and computer vision, with a main focus on:
- Inverse problems applied to signal and image restoration, reconstruction, analysis and processing,
- High resolution, high framerate imaging systems,
- Dynamic imaging,
- Machine learning applied to image processing and analysis
MINDS research activities in computational imaging are centered on medical imaging, with a main focus on ultrasound. They concern the whole image formation and interpretation process including raw data acquisition by individual sensors using conventional and non-traditional sampling schemes, beamforming, image reconstruction, processing and analysis. The main research topics in computational ultrasound cover, but are not limited to:
- Inverse problems: signal and image restoration, resolution enhancement, beamforming, compressed sensing, tissue motion estimation, flow velocity estimation
- Linear and non-linear ultrasound imaging
- Various clinical applications: oncology, cardiology, gynecology
MINDS also conducts research on other imaging modalities:
- Computed tomography applied to endodontics
- Multi-modal (ultrasound – magnetic resonance imaging) image registration and fusion applied to gynecology.
MINDS research activities in computer vision benefit from a broad range of expertise in camera calibration, stereo vision and motion analysis.
The main current research topics in computer vision include:
- Still image analysis: interactive segmentation, over-segmentation, object identification;
- Temporal image series: change detection, re-identification;
- Video analysis: action recognition, scene reconstruction.
- Photographic landscape observatories;
- Security: video analysis for criminal investigation, public transportation.
MINDS also considers other applications like fish ecology and astronomical image processing.