节点文献
利用立体视觉的砂轮地貌三维表面重构
3-D surface reconstruction of monolayer grinding wheel topography based on stereo vision
【摘要】 提出一种利用双目图像以恢复砂轮地貌三维表面的方法,通过基于灰度的面匹配法对双目图像精确匹配,然后利用双目视觉成像原理,求出该样本点在世界坐标系的坐标。在此一系列样本点的世界坐标基础上,可得待恢复区域的三维外轮廓线和外轮廓面,通过面元内插法可以得到三维场景的精确恢复。实验结果表明,该算法能有效恢复砂轮地貌的三维表面信息。
【Abstract】 A method to reconstruct 3D surface of monolayer grinding wheel topography based on stereo vision is presented. Through precise matching on binocular images with module match method according to grey information , the coordinates in World Coordinate System of sample-points are obtained based on the principle of stereo vision. The contour line and skeleton map are obtained with those coordinates and then a 3D surface is reconstructed accurately by using the surface interpolation method. Experimental result shows that the method can effectively obtain 3D surface of monolayer grinding wheel topography.
【Key words】 Stereo vision; Topography of grinding wheel; 3D reconstruction; Match;
- 【文献出处】 光电工程 ,Opto-electronic Engineering , 编辑部邮箱 ,2005年02期
- 【分类号】TP391.4
- 【被引频次】12
- 【下载频次】225