摘要
目前未知对象的自动三维重建已经成为计算机视觉和机器人视觉研究的热点之一。在未知对象自动三维重建过程中除了视觉规划、数据获取、数据配准等技术之外,如何自动判别出继续或终止三维重建过程也显得尤为重要。提出了一种基于高斯定理的自动重建过程中的自终止判别方法。该判别方法主要利用相邻两次获取数据所围体积的相对变化量作为判别准则,如果该变化量小于预先给定的阈值,那么则终止数据获取;否则继续数据获取。实验证实了该方法可行性。
Automatic 3D reconstruction of 3D unknown objects has attracted much attention from computer vision robotics and vision researches. Except for how to find a suitably sensor poses, configurations the sensor position, it's very important that when to terminate the process of reconstruction or inspection 3D objects. A self-termination criterion is proposed for judging the completion condition in the measurement and reconstruction process. The termination condition is derived based on changes in the volume obtained from the last two successive viewpoints. Experimental results are given in the digitization and reconstruction of freeform objects using our self- termination judgment method.
出处
《工程图学学报》
CSCD
北大核心
2007年第5期76-81,共6页
Journal of Engineering Graphics
基金
国家自然科学基金资助项目(50605007)
福建省自然科学基金计划资助项目(E0610013)
关键词
计算机应用
自终止
高斯定理
三维重建
视觉规划
computer application
self-termination
Gauss's theorem
3D reconstruction
view planning