Dissertation
Dissertation > Industrial Technology > Automation technology,computer technology > Computing technology,computer technology > Computer applications > Information processing (information processing) > Pattern Recognition and devices > Image recognition device

Research on Camera Localization Based on Monocular Vision

Author ShenHuiJie
Tutor WangRongBen
School Jilin University
Course Vehicle Operation Engineering
Keywords Camera positioning Camera Calibration Monocular vision ranging Image Stitching Image registration Image Fusion
CLC TP391.41
Type Master's thesis
Year 2009
Downloads 535
Quotes 9
Download Dissertation

Computer vision because it contains a wealth of environmental information has been widespread concern , one of the monocular vision - based positioning technology is the computer vision research direction . General steps based on monocular vision camera positioning method first pass between the image and the template matching relationship corresponding points to create a constraint homography , then homography solving the camera position parameters . However , the corresponding points between images , and the corresponding points between the image and the template matching has been the difficulty of the image processing and analysis , the matching correctness directly affect the positioning accuracy . Therefore, the positioning process to minimize or avoid matching is very practical value . In view of the above problems , this paper two camera positioning : positioning method ( triangle method ) based on the geometric relationship between the camera and the camera positioning method based on image stitching . Ranging monocular vision model based on the geometric relationship between the camera positioning method , establish the geometric relationship between the camera and the scene known feature points can be derived through a series of triangular camera dimensional coordinate plane position , the method to avoid the positioning process on the image point matching problem , the continuous positioning of the camera can be achieved . Camera positioning method based on the image stitching and its essence is use of the affine transformation of the image splicing process , i.e. after the plurality of pairs of matching points in the two adjacent positions in the image is found , solving the affine transformation model , to thereby obtain two images transformation matrix and translation matrix , and launch the camera position changes , the initial camera positioning . At the same time, the image of the scene to get a larger field of view image stitching , more environmental information . The experiments show that the two positioning methods of this study is feasible , there is a certain degree of robustness , initially completed a monocular vision camera positioning method based exploration and research , for the realization of the lunar rover intelligent vehicle based visual track speculate and tracking provide the necessary technical support .

Related Dissertations
More Dissertations