Image Processing Group
We are interested in the development of models and algorithms for image processing and computer vision.
Our main lines of work are focused in
a) 3D reconstruction and view synthesis for sport events and 3D Television and cinema
- Depth computation and 3D reconstruction, shape from silhouette, new view synthesis, inpainting methods for new view synthesis, processing of data for 3D Television.
b) Development of tools for digital video post-production
- Image and video denoising, super-resolution, video editing, image and video inpainting, segmentation, tracking, motion estimation, de-interlacing, video format conversion, film post-production algorithms, level lines selection for segmentation, compression and coding.
c) Object identification including logo detection.
- Affine invariant object recognition, logo detection.
d) Color image processing and digital photography
- Image fusion, perceptual-based color enhancement, tone mapping, HDR imaging techniques.
e) Variational models for image processing and computer vision
- Development of numerical methods and optimization methods for optical flow and depth computation and its GPU implementation.
f) Applied Mathematics: geometric partial differential equations
- Flux limited diffusion equations, interface motion by curvature and variational principles for transport and irrigation networks.