Martin Čadík: Computational photography

16.50 - 17.15


Contemporary cameras not only capture the image, but also count. Computational photography methods use computer graphics, image processing, and computer vision algorithms to expand the possibilities of both classical and digital photography. In the lecture I will present the most interesting methods of computational photography, which allow, for example, to extend the depth of field of the image, increase the exposure flexibility of the camera, or reduce motion blur.

I will also introduce visual localization, which aims to estimate the position of the photographer from the photography itself, with a number of unconventional applications.