CASIA OpenIR  > 毕业生  > 硕士学位论文
Thesis Advisor马颂德
Degree Grantor中国科学院自动化研究所
Place of Conferral中国科学院自动化研究所
Degree Discipline模式识别与智能系统
Abstract基元提取是Marr的视觉计算理论的重要问题。早期视觉主要关注于如何更 有效、更鲁棒的提取各种基元。除了点和直线外,曲线也是很重要的基元。它 们比点和直线要含有更丰富的信息。本文提出了一种有效的提取参数曲线基元 的方法。该方法对直线和参数曲面同样适用。 参数曲线和曲面可以用来解决许多视觉问题,尤其是基于模型的视觉问 题,如方位估计,三维重建,建模,物体识别,机器人定位。这是因为模型视 觉的基本假设是:周围环境由几何模型已知的物体组成。组成这些物体的表面 以及物体的轮廓线都是参数曲线和曲面。本文提出了一种基于物体表面上平面 曲线特征的方位估计方法。该方法为线性方法并且适用于一般的多项式曲线 (≥2)。本文还提出了从三幅及三幅以上透视图中的椭球轮廓线重建椭球的 方法。我们证明如果摄像机经过定标,三幅图像的轮廓线包含所有椭球及其方 位的信息:如果摄像机未经过定标,我们可以得到椭球的射影重建结果。这些 算法都是线性的。我们相信将我们的方法与文献中已有方法相接合可以为工业 应用建立实用的视觉系统。
Other AbstractPrimitive extraction is fundamental in Marr's vision paradigm. Early vision is largely concerned about how to efficiently and robust extract various kinds of primitives. Apart from points and lines, curves are also very important primitives. They contain much more information than points and lines. In this thesis, we propose an efficient method for the extraction of parametric curved primitives. This method also works well for lines and parametric surfaces. Parametric curved primitives can be used to solve many vision problems, especially problems in model-based vision. This is due to the general assumption in model-based vision that the environment consists of manufactured objects whose geometry is known a priori. The surfaces of these objects and the outline of these surfaces can all be modeled as parametric surfaces and curves. In this thesis, a pose determination method based on high degree planar curved features on the surface of an object is presented. This method is linear and generally applicable to any high degree (≥> 2) planar curved features. A method to reconstruct ellipsoid from its occluding contours observed in three images is also presented. The most prominent feature of this method is that the relationship of an ellipsoid and its perspective projections is linear. The reconstruction algorithm works for both the camera calibration case and the weak calibration case. In the case of weak calibration, the reconstructed ellipsoid is up to a projectivity. We believe that using our methods, together with the existing methods in the literature, practical vision system can be built.
Other Identifier388
Document Type学位论文
Recommended Citation
GB/T 7714
李利. 基于曲线特征的物体定位和三维重建[D]. 中国科学院自动化研究所. 中国科学院自动化研究所,1996.
Files in This Item:
There are no files associated with this item.
Related Services
Recommend this item
Usage statistics
Export to Endnote
Google Scholar
Similar articles in Google Scholar
[李利]'s Articles
Baidu academic
Similar articles in Baidu academic
[李利]'s Articles
Bing Scholar
Similar articles in Bing Scholar
[李利]'s Articles
Terms of Use
No data!
Social Bookmark/Share
All comments (0)
No comment.

Items in the repository are protected by copyright, with all rights reserved, unless otherwise indicated.