摘要
针对普通数码相机拍摄的序列图像,提出了一种快速的交互式三维重建方法。首先基于SFM技术得到了相机的内外部参数;然后在图像上交互取点以构建逼近物体轮廓的多边形,结合该多边形,利用轮廓跟踪算法提取出精确的外形轮廓;最后通过匹配轮廓点的方式重建出三维模型。实验结果表明该重建方法是快速而有效的。
A rapid interactive 3D Reconstruction,which aimed at image sequences of objects that are taken by common digital camera, was given in this paper. First some useful parameters of camera were obtained by SFM. And then user trace out the boundary of the object with polygons on the image by picking some points. By analyzing these polygons, contour tracking algorithm was used to determine the precise boundaries of objects. Finally, use the automatic matching algorithm to get correspond points. The 3D model of the object can be reconstructed with the correspond points. Experimental results show that the reconstruction method is rapid, valid and efficient.
出处
《微处理机》
2012年第4期49-52,共4页
Microprocessors
关键词
计算机视觉
三维重建
SFM算法
Computer Vision
3 D Reconstruction
Structure From Motion Algorithm
