摘要
To solve the problem of camera spatial location based on mono vision, a new algorithm is proposed: relative pose parameters between the camera and the marked point are found by the spatial geometry information of the special marked point; then inferring the camera's spatial location coordinates by it. Instead of complicated nonlinear calculation to each frame, the algorithm gets the initial value for consecutive pose calculation by a particular pose between camera and the marked object, then doing calculation of pose tracking to the following frame by the restriction between frames. It is proved that the algorithm can not only guarantee the precision but also reduce the amount of calculation in the experimental and practical applications.
- 出版日期2008
- 单位西北工业大学