摘要
以投影几何学以及双目立体视觉原理为理论基础,对移动机器人的三维重建技术进行研究,对移动机器人漫道过程中所在的兴趣区域的场景进行较为精确的建模.设计了机器人的快速建模方法,利用迭代最近点算法(ICP),完成了多个局部三维场景模型的融合.同时,结合栅格投射理论,完成了对全局三维场景模型的更新.利用栅格模型重建的三维场景,具有环境信息丰富,模型描述精确的特点,可以应用于移动机器人导航领域.
Based on the theories of stereo vision and projective geometry, we study the technology of 3D reconstruction system, and build an accurate model for the interested area of a mobile robot during it roaming the road, By using the iterative closest point algorithm (ICP), we design a robot fast-modeling method which can integrate a number of local 3D scene models. At the same time, by cornbining grid projection theory, we'accomplish the updating of global 3D scene model. Reconstructed by binocular stereo vision, the 3D scene with characteristics of abundant environment information and accurately described model can be applied for the navigation of mobile robot
出处
《小型微型计算机系统》
CSCD
北大核心
2012年第4期873-877,共5页
Journal of Chinese Computer Systems
基金
辽宁省自然科学基金项目(20092006)资助
关键词
移动机器人
三维场景重建
双目立体视觉
迭代最近点
栅格投射
mobile robot
3D scene reconstruction
binocular stereo vision
iterative closest point
evidence grid projection