作者：李小菁1,2, 黄振宇1,2, 凌珊1,2
作者单位：1. 广东省现代几何与力学计量技术重点实验室, 广东 广州 510405;
2. 广东省计量科学研究院, 广东 广州 510405
关键词：双目视觉;动态三维测量;光栅结构光;光条匹配
摘要：
基于矩形光栅结构光和双目立体视觉测量技术，提出一种物体表面三维形貌的动态测量方法。采用基于光平面引导的稠密光栅光条匹配方法完成左右摄像机光条图像的匹配，该方法先通过各光平面对应的单应矩阵实现左右图像点的对应，再结合极限约束及空间点距离约束完成左右光条中心点的全局最优匹配，以此解决由两摄像机和投影仪组成光栅结构光测量系统中光条匹配难题，对匹配完成的图像光条点应用双目立体视觉测量原理完成物体表面的三维重建。实物实验中，对姿态变化的手臂进行测量，结果证明仅需投影单幅光栅图案可实现手臂三维形貌的测量，验证该方法的有效性。
Dynamic 3D measurement method based on raster binocular stereo vision sensor
LI Xiaojing1,2, HUANG Zhenyu1,2, LING Shan1,2
1. Guangdong Provincial Key Laboratory of Modern Geometric and Mechanical Metrology Technology, Guangzhou 510405, China;
2. Guangdong Provincial Institute of Metrology, Guangzhou 510405, China
Abstract: Based on the rectangular grating structure light and binocular vision measuring technology, a method for dynamically measuring 3D topography on the object surface is proposed. A matching method of high-density grating light stripes based on the guidance of light plane is adopted to complete matching of the left and right camera light-strip images, in which correspondence of the left and right image points can be firstly realized by homography matrix corresponding to each light plane, and then the optimal overall matching of the center points on the left and right light strips can be achieved by combining the limit constraint and the space point distance constraint. As a result, the matching problem in the grating structured light measuring system that is composed of two cameras and projectors is solved. Finally, the 3D reconstruction on the object surface is completed according to the binocular stereo vision measuring principles. In the physical experiment, the measurement of a moving arm is carried out, and the results prove that it only needs to project a single grating image to measure its 3D topography, thus, effectiveness of such method is verified.
Keywords: stereo vision;dynamic 3D measurement;raster structure light;light matching
2018, 44(9): 35-38,50 收稿日期: 2018-02-05;收到修改稿日期: 2018-03-12
基金项目:
作者简介: 李小菁(1991-),女,广东韶关市人,硕士,主要从事视觉测量及几何量计量检测技术研究
参考文献
[1] SHIRMOHAMMADI S, FERRERO A. Camera as the instrument:the rising trend of vision based measurement[J]. IEEE Transactions on Instrument Measure, 2014, 17(3):41-47
[2] 华才健, 蹇红梅, 张燕. 基于机器视觉的火花塞间隙检测技术[J]. 中国测试, 2014, 40(5):108-110
[3] 郭继平, 李阿蒙, 于冀平, 等. 双目立体视觉动态角度测量方法[J]. 中国测试, 2015, 41(7):21-23
[4] DU Q J, SHI X, DAI B, et al. Binocular stereo vision system for a humanoid robot[J]. International Journal of Computer Applications in Technology, 2013, 46(4):316-322
[5] LOHRY W, ZHANG S. Fourier transform profilometry using a binary area modulation technique[J]. Optical Engineering, 2012, 51(11):3602-3609
[6] 毛先富, 苏显渝, 陈文静, 等. 改进傅里叶变换轮廓术的测量算法研究[J]. 光学学报, 2008, 28(7):1291-1295
[7] 范剑英, 付瑶, 王洋, 等. 双三步三角形相移编码结构光三维测量技术[J]. 黑龙江大学自然科学学报, 2016, 33(2):251-255
[8] FU Y J, LUO Q. Fringe projection profilometry based on a novel phase shift method[J]. Optics Express, 2011, 19(22):21739-21747
[9] 韦争亮, 钟约先, 袁朝龙. 基于彩色栅线的结构光动态三维测量技术研究[J]. 光学技术, 2009, 35(4):569-574
[10] JANG W, JE C, SEO Y, et al. Structured-light stereo:Comparative analysis and integration of structured-light and active stereo for measuring dynamic shape[J]. Optics and Lasers in Engineering, 2013, 51(11):1255-1264
[11] WONG A K C, NIU P Y, HE X. Fast acquisition of dense depth data by a new structured light scheme[J]. Computer Vision and Image Understanding, 2005, 98(3):398-422
[12] 张广军. 视觉测量[M]. 北京:科学出版社, 2008:26-40.
[13] STEGER C. An unbiased detector of curvilinear structures[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 1998, 20(2):113-125
[14] 魏振忠, 樊妍睿, 侯亚荣. 光栅式双目立体视觉传感器光条快速匹配算法[J]. 北京航空航天大学学报, 2012, 38(5):591-594
[15] ZHANG Z Y. A flexible new technique for camera calibration[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2000, 22(11):1330-1334
