Mathematical theory of image straightening with applications to camera calibration

Extracting quantitative measures from roadway camera images requires calibration of the camera. Roadway features such as fog lines and lane boundaries can provide the information necessary to calibrate the cameras on the fly. In addition, "straightening" the image allows for linear measurements in the image which in turn allows the use of linear correlation to measure features and traffic speed. This paper presents the theory for straightening and camera calibration