Computer Engineering and Applications ›› 2012, Vol. 48 ›› Issue (6): 205-208.

• 图形、图像、模式识别 • Previous Articles     Next Articles

3D reconstruction method based on image silhouette

CAO Yu1,2, CHEN Xiuhong2   

  1. 1.School of Internet of Things, School of Digital Media, Jiangnan University, Wuxi, Jiangsu 214122, China
    2.School of Digital Media, Jiangnan University, Wuxi, Jiangsu 214122, China
  • Received:1900-01-01 Revised:1900-01-01 Online:2012-02-21 Published:2012-02-21

基于图像轮廓的三维重建方法

曹 煜1,2,陈秀宏2   

  1. 1.江南大学 数字媒体学院 物联网工程学院,江苏 无锡 214122
    2.江南大学 数字媒体学院,江苏 无锡 214122

Abstract: According to the plane mirror imaging principle ,a method which based on the outline of the silhouette of a single image to create the visual hull is proposed. The method uses angled plane mirror system to simulate multi camera shooting at the same time, it can get the corresponding camera parameters to realize the 3D reconstruction of the object by analyzing the epipolar geometry relationship between each silhouette. Experimental results show that the method is simple and effective, and can reconstruct the 3D shape of the target object without the calibration result obtained through special laboratory instrument. And the method has extensive practical value.

摘要: 根据平面镜成像原理,提出了一种基于单幅图像的侧影轮廓线可见外壳重建方法。利用成角度平面镜装置模拟实现多相机同时拍摄,对图像中各轮廓线之间极线几何关系的分析得到对应的相机参数,实现目标物体的三维重建。实验结果表明,该方法简单有效,不需要通过实验室条件下的特殊仪器进行标定即可实现目标物体的三维重建,具有一定的实用价值。