System for Automatic Camera Calibration Robust Against Blur and Lighting Conditions Changes

This paper presents a system for precise intrinsic and extrinsic camera calibration for monocular and stereo cameras. The presented approach is based on some well established research in the field and utilizes calibration patterns. The novelty of the presented method is the automatic detection of motion blur, that is common to the process of calibration based on moving the camera or the calibration object, especially under bad lighting conditions. The method allows for sub-pixel re-projection precision for cameras with perspective lenses and can be extended to omnidirectional cameras as well.