Category:Geometry in computer vision

From Wikipedia, the free encyclopedia

Geometry in computer vision is a sub-field within computer vision dealing with geometric relations between the 3D world and its projection into 2D image, typically by means of a pinhole camera. Common problems in this field relate to

  • Reconstruction of geometric structures (for example, points or lines) in the 3D world based on measurements in 2D images
  • Matching constraints which are satisfied for 2D structures (for example, points or lines) in two or more images of the same scene when these structures correspond in the 3D scene.

The geometric structures studied in this field does not have to be restricted to points or lines in two or three dimensions but can also be related to entire objects, for example the pose of such an object.