Sara, Radim2015-02-212015-02-2120071467-8659https://doi.org/10.1111/j.1467-8659.2007.01045.xComputer Vision is a discipline whose ultimate goal is to interpret optical images of real scenes. It is well understood that such a problem is cursed by ambiguity of interpretation and uncertainty of evidence. Despite imperfectness of results due to the scenes never following our prior models exactly, Computer Vision has achieved a significant progress in the past two decades.This talk will outline the quest of 3D Computer Vision by describing a processing pipeline that receives a heap of unorganized images from unknown cameras and produces a consistent 3D geometric model together with camera calibrations. We will see how new algorithms allow the standard conception of the pipeline as a series of independent processing steps gradually transform to a single complex, yet efficient vision task. We will identify some points where linking Computer Vision and Computer Graphics would bring significant progress.10.1111/j.1467-8659.2007.01045.xxix-xix