摘要
为实现形体的快速三维重建,在分析现有各类重建算法的基础上,针对不规则形体给出了一种三维表面点的重建方法。首先进行摄像机自定标,求取相机的内、外参数,然后对采集的图像进行处理,得到物体的轮廓,最后使用地毯式的搜索算法搜索并重建了物体的表面点。分析了不同搜索步长情况下算法的执行效率,在综合考虑逼真度和执行速度的情况下,给出了优化的步长选择,为表面的三角网格化和纹理恢复提供了依据。
A three dimensional surface-point reconstruction method suitable for irregular object is given on the basis of analyzing the current image-based 3D reconstruction methods in order to realize fast modeling of objects. Firstly, realize the camera self-calL brafion and get the intrinsic and extrinsic parameters of the camera. Secondly, the contour of objects is calculated by collected image processing. Finally, the carpet-search algorithm is used to search and reconstruct the surface points of objects. The effect of this algorithm at different search steps was analyzed and an optimal step-selection was given by comprehensive consideration of fidelity and execution speed, which provide some bases for triangle-grid and texture recovery of the surface
出处
《东北林业大学学报》
CAS
CSCD
北大核心
2008年第8期54-57,共4页
Journal of Northeast Forestry University
关键词
不规则形体
三维重建
摄像机自定标
表面点
Irregular objects
3D reconstruction
Camera self-calibration
Surface points