(An) active camera calibration using vanishing points = 소실점을 이용한 능동적인 카메라 캘리브레이션

Cited 0 time in webofscience Cited 0 time in scopus
  • Hit : 389
  • Download : 0
This dissertation deals with the issues for the automation of the camera calibration process for robotic applications. The applications may include mobile robot localization and 3-D object modeling, where autonomy is essential. For example, a mobile robot needs a constant relocalization during navigation, the reliability of which, in large part, is influenced by unknown systematic errors in the postulated sensor model. In this dissertation, camera rotation and focal length variation are assumed to be the primary sources that cause systematic errors in 3-D measurements by the monocular vision system. In order to cope with such changes in camera parameters, an active calibration method is proposed in which vanishing points are primary tools to realize the aim. The proposed calibration method consists of two fundamental stages: unsupervised line extraction for vanishing point detection and estimation of the camera parameters. Unsupervised line clustering is a fundamental step to extract vanishing points from parallel lines. It is assumed that each line support region (LSR) in an image is composed of pixels that share similar gradient orientation values. Therefore, by an appropriate partitioning of gradient space, the sets of parallel lines can be more easily extracted. Bhattacharyya distance is introduced to define a measure for cluster separability and thereafter to validate the number of inherent clusters. Subsequent to the cluster validation stage, each extracted line support region undergoes a consistency test to ensure its validity in terms of uncertainty descriptors. For the consistency test, an entropy-based line selection scheme is formulated and a theory from robust statistics is adopted. The feasibility of the proposed method is assessed by comparing the estimated positions of vanishing points with manually detected ones. The calibration procedures exploit several properties from projective geometry, mainly from the basic property of geometric invarian...
Chung, Myung-Jinresearcher정명진researcher
한국과학기술원 : 전기및전자공학과,
Issue Date
150991/325007 / 000945258

학위논문(박사) - 한국과학기술원 : 전기및전자공학과, 1999.2, [ [x], 113 p. ]


Unsupervised line clustering; Projective geometry; Vanishing point; Camera calibration; Robust statistics; 진화 연산; 자율적인 선분 클러스터링; 사영 기하학; 소실점; 카메라 캘리브레이션; Evolutionary programming

Appears in Collection
Files in This Item
There are no files associated with this item.


  • mendeley


rss_1.0 rss_2.0 atom_1.0