科研之友
微信
新浪微博
Facebook
分享链接
摘要
On the theoretical analysis, the six factors relating to the measuring precision of a monocular vision pose measurement system with three corresponding feature points were investigated. These factors are the distance between marker and camera, size of the marker, quantitative errors of camera, calibration errors of intrinsic parameters of camera, position errors of feature points, and the error introduced by pose estimation algorithm. The simulations and imaging experiments were performed to verify the validity of the theoretical analysis. It shows that the most effective factor is the measuring distance. The longer focal length and smaller size of a pixel on CCD can reduce the quantitative error, and the bigger error size of target can improve the precision of feature extraction and can minimize the impaction of extraction error caused by image processing algorithm. The noise introduced by electronic hardware is also an important factor that can not be ignored during the system designing phase. The focal length, position of intersection of optical axis and the image plane on CCD, aberrations of lenses of the camera should be calibrated after the camera is manufactured during the calibration phase, and the error of calibration is the second significant factor for the whole measurement system. The error introduced by pose estimation algorithm can be ignored because it is much smaller than the whole errors. The position errors of feature points extracted from the image is the main impaction of precision loss of system.
- 出版日期2014
- 单位中国科学院大学
