Principal point determination for camera calibration
Date of Award
Ph.D. in Electrical and Computer Engineering
Department of Electrical and Computer Engineering
Advisor: John S. Loomis
Calibration is a fundamental task in computer vision, and it is critical for a growing number of applications. Identification of the principal point parameter is a significant step for calibration, because its accuracy will strongly effect the other parameters and the overall accuracy of the calibration. Additionally, some camera properties require slightly different definitions for the principal point. This work explores several novel techniques for highly accurate principal point estimation, all of which utilize simple planar checkerboard targets. First, an improved and automated corner detection algorithm is described. Checkerboard image corner points are located as saddle points, and the Hough transform is applied to remove spurious points and group them into rows and columns. The lines formed from these rows and columns are used to identify vanishing points. Multiple vanishing points lie along horizon lines, and two different techniques based on horizon lines are described for estimating the principal point. It is also possible to identify the principal point using images of a pair of checkerboards, one behind the other, that are nominally perpendicular to the camera's optical axis. This problem requires additional corner-point processing to separate the two checkerboards in the images, and corrections are developed to handle orientation errors such as small rotations and translations between the checkerboards and image plane. Experimental results for these methods are presented and their accuracy and possible applications are discussed.
Cameras Calibration, Focal planes, Computer vision, Computer Science, Electrical Engineering, Camera Calibration, Computer Vision, Corner detection, hough transform, principal point
Copyright 2017, author
Alturki, Abdulrahman Saleh, "Principal point determination for camera calibration" (2017). Graduate Theses and Dissertations. 1268.