基于面结构光几何关系的三维数字化研究  

3D digitization based on geometry of surface structured light

作　　者：杨海清[1] 王洋洋 Yang Haiqing;Wang Yangyang(College of Information Engineering,Zhejiang University of Technology,Hangzhou 310023,China)

机构地区：[1]浙江工业大学信息工程学院,杭州310023

出　　处：《计算机应用研究》2018年第7期2237-2240,共4页Application Research of Computers

基　　金：浙江省自然科学基金资助项目(LY13F010008);浙江省科学计划资助项目(2015F50009)

摘　　要：为降低颜色串扰对提取结构光图像特征的影响,通常将多像素条纹细化为单像素线条,降低了重建图像空间分辨率。针对这一问题,设计一个基于面结构光几何关系的三维数字化系统,并提出一种以面结构光投影仪像素为单位的特征提取与匹配新方法。首先,投影仪向被测物体投射设计好的面结构光,为被测物体增加有效特征,双目摄像机捕获被测物体两张图像,设置阈值滤除图像背景、去除噪声,提高图像质量;然后,基于单位投影像素提取特征点,充分利用投影图像的全像素空间,使得特征点采样密度提高,根据特征点间的几何关系进行匹配,提高计算效率;最后,根据三角测量原理将特征点图像坐标转换成世界坐标,形成高密度点云;在Geomagic Studio 2013平台上对被测物体进行三维重建,获得高质量的三维重建模型,实验结果达到了预期效果。In order to extract image features under structured-light projection,multi-pixel fringes are usually refined into single-pixel due to color crosstalk of light,which results in low spatial resolution of image reconstruction. To solve the problem,this paper designed a 3 D-digitization system based on geometry of surface structured light,meanwhile proposed a new method of feature extraction and matching based on projector＇s pixel unit of surface structured light. First of all,this paper projected surface structured light with a specific mode of texture on an object for the purpose of forming image features on the surface of the object. It used a binocular camera to capture two images of the object,and removed these images noise through the threshold filtering method. Then,this paper extracted feature points based on the unit of projector pixel and matched according to the geometric relationship in each image and enhanced the computational efficiency. Finally,this paper transformed the image coordinates of the matched feature points into the world coordinates through the method of triangulation,which resulted in a point cloud with high density. The experiment obtains a high-resolution 3 D reconstruction model of the object and achieves expectation of the research on the Geomagic Studio 2013.

关 键 词：特征点提取 特征点匹配 结构光测量 点云 三维数字化 

分 类 号：TP391.4[自动化与计算机技术—计算机应用技术]