• CN:11-2187/TH
  • ISSN:0577-6686

›› 2014, Vol. 50 ›› Issue (2): 42-47.

• 论文 • 上一篇    下一篇

基于单幅灰度图像的快速三维重建方法研究

高欣健;张旭东;高亚捷;Peter Ott   

  1. 合肥工业大学计算机与信息学院;德国海尔布隆大学光学实验室
  • 发布日期:2014-01-20

Fast Method of 3D Surface Reconstruction Based on Single Gray Image

GAO Xinjian;ZHANG Xudong;GAO Yajie;OTT Peter   

  1. School of Computer and Information, Hefei University of Technology Optical Engineering Lab., Heilbronn University
  • Published:2014-01-20

摘要: 对含有阴影的物体表面快速三维重建问题,提出一种融合阴影恢复形状(Shape from shading, SFS)数据与旋转对称激光三角传感器数据的基于单幅灰度图像的快速智能测量方法。针对SFS方法重建过程中出现的二义性问题,利用旋转对称激光三角传感器的关键数据对其进行修正。同时针对旋转对称激光三角传感器对未知物体表面需要逐点测量,测量速度慢的问题,利用修正后的SFS重建结果对旋转对称激光三角传感器进行智能视觉引导,对含有阴影的物体表面梯度变化较大的点及特征点进行精确测量。在此基础上搭建集成SFS方法的激光三角传感器高精度移动测量平台系统。该系统根据物体表面梯度变化的特点自动选取阈值调节测量精度,从而快速、精确地完成对未知物体表面的三维重建。并通过试验验证了方法的可行性和有效性。

关键词: 三维重建;数据融合;激光三角

Abstract: Fast method of 3D surface reconstruction is purposed for the problem of surface that contains shadow. The data from shape from shading and rotational symmetric triangulation sensor based on single gray image are be integrated. Purposed for the problem of ensure two meanings of the data of shape from shading, the key data from rotational symmetric triangulation sensor are be used. Purposed for the problem that the rotational symmetric triangulation has to measure the surface point by point and the low measurement speed, the corrected data from Shape from shading are be used to guide the rotational symmetric triangulation sensor to measure the points of the surface that contains shadow which have great gradient and the feature points precisely. A system integrates data of shape from shading and data of rotational symmetric triangulation sensor is built. The accuracy of measurement of this system will be changed by the gradient of the surface. Therefor the surface that contains shadow can be reconstructed fast and precisely. Experiments results show that the method is with higher accuracy and robust.

Key words: 3D reconstruction;data integrating;laser triangulation

中图分类号: