摘要
计算机视觉领域的三维重建主要指从二维图像还原出物体的三维结构。本文主要介绍了单目三维重建技术,即使用单个摄像机所成图像,寻找图像的对应特征,根据这些特征的变化建模,逆向推导出物体的三维信息。本文主要从单目三维重建较为典型的单目三维重建算法包括明暗法、光度立体视觉法、轮廓法等来展现这种对应特征的建模问题,并介绍了这些算法的适用情况及不足。
In the field of computer vision,three-dimensional reconstruction mainly refers to restoring the three-dimensional structure of objects from two-dimensional images.This paper mainly introduces the technique of monocular 3D reconstruction,that is,using the image produced by a single camera,looking for the corresponding features of the image,modeling based on the changes of these features,and reversely deducing the 3D information of the object.In this paper,the typical monocular 3D reconstruction algorithms including light and shade method,photometric stereo method,contour method,etc.are mainly used to showthe modeling problems of corresponding features,and the application and shortcomings of these algorithms are introduced.
作者
包晓敏
白晨
BAO Xiaomin;BAI Chen(Zhejiang Sci-Tech University,Hangzhou 310018,China)
出处
《智能计算机与应用》
2020年第7期49-52,共4页
Intelligent Computer and Applications
基金
浙江省重点研发计划项目(2018C01133)
关键词
三维重建
单目
对应特征
建模
3D reconstruction
monocular
corresponding features
modeling
