测绘学报 ›› 2018, Vol. 47 ›› Issue (5): 631-643.doi: 10.11947/j.AGCS.2018.20170365

• 摄影测量学与遥感 • 上一篇    下一篇

互联网众源照片的三维重建定位技术

袁一1,4, 程亮1,2,3,4, 宗雯雯1,4, 李舒怡1,4, 李满春1,2,4,5   

  1. 1. 南京大学江苏省地理信息技术重点实验室, 江苏 南京 210023;
    2. 南京大学中国南海研究协同创新中心, 江苏 南京 210023;
    3. 南京大学软件新技术与产业化协同创新中心, 江苏 南京 210023;
    4. 南京大学地理与海洋科学学院, 江苏 南京 210023;
    5. 南京师范大学江苏省地理信息资源开发与利用协同创新中心, 江苏 南京 210023
  • 收稿日期:2017-06-28 修回日期:2018-03-01 出版日期:2018-05-20 发布日期:2018-06-01
  • 通讯作者: 程亮 E-mail:lcheng@nju.edu.cn
  • 作者简介:袁一(1993-),女,硕士生,研究方向为激光雷达与摄影测量数据集成处理与应用。E-mail:yuanyi_1213@163.com
  • 基金资助:
    国家重点研发计划(2017YFB0504205);国家自然科学基金(41622109;41371017)

Crowd-sourced Pictures Geo-localization Method Based on 3D Reconstruction

YUAN Yi1,4, CHENG Liang1,2,3,4, ZONG Wenwen1,4, LI Shuyi1,4, LI Manchun1,2,4,5   

  1. 1. Jiangsu Provincial Key Laboratory of Geographic Information Science and Technology, Nanjing University, Nanjing 210023, China;
    2. Collaborative Innovation Center for the South Sea Studies, Nanjing University, Nanjing 210023, China;
    3. Collaborative Innovation Center of Novel Software Technology and Industrialization, Nanjing University, Nanjing 210023, China;
    4. School of Geographic and Oceanographic Sciences, Nanjing University, Nanjing 210023, China;
    5. Jiangsu Center for Collaborative Innovation in Geographical Information Resource Development and Application, Nanjing Normal University, Nanjing 210023, China
  • Received:2017-06-28 Revised:2018-03-01 Online:2018-05-20 Published:2018-06-01
  • Supported by:
    The National Key Research and Development Program (No.2017YFB0504205);The National Natural Science Foundation of China (Nos.41622109;41371017)

摘要: 随着电子产品的普及和互联网发展,人们习惯性地将照片上传到主流图片分享网站和社交媒体。这些照片很多都没有地理坐标或者仅有模糊的位置信息。如果互联网上任意来源的电子照片能被恢复出真实地理位置,它们将可以应用在城市建设、城市景观分析和犯罪追踪等很多方面。本文提出了一种互联网众筹照片的三维重建定位技术,该技术将结构化组织的街景数据作为参考数据集,使用三步走策略:图像检索粗定位、图像匹配细筛选和三维重建精定位,来给不明来源的众筹照片附上精确的地理标签。本文通过摄影测量原理恢复待查询照片周围的三维空间信息,较之前Zamir和Shah的方法,定位精度中值从256.7 m提升到69.0 m,平均值从350.4 m提升到206.0 m,在50 m精度要求下的照片数量占比从17.2%提升到43.2%。另一个发现是在重建误差成因方面,待查询照片的相机距离主拍摄目标越近,定位总误差越小。本文所提出的照片定位技术提供了灵活的参数,使其应用范围不仅仅局限于小区域,也可以扩大至城市、国家尺度。

关键词: 街景数据, 照片定位, 图像检索, 三维重建, 精度分析

Abstract: People are increasingly becoming accustomed to taking photos of everyday life in modern cities and uploading them on major photo-sharing social media sites. These sites contain numerous pictures, but many have incomplete or blurred location information. The geo-localization of crowd-sourced pictures enriches the information contained therein, and is applicable to activities such as urban construction, urban landscape analysis, and crime tracking. However, geo-localization faces huge technical challenges. This paper proposes a method for large-scale geo-localization of crowd-sourced pictures. Our approach uses structured, organized Street View images as a reference dataset and employs a three-step strategy of coarse geo-localization by image retrieval, selecting reliable matches by image registration, and fine geo-localization by 3D reconstruction to attach geographic tags to pictures from unidentified sources. 3D reconstruction based on close-range photogrammetry is used to restore the 3D geographical information of the crowd-sourced pictures, resulting in the proposed method improving the median error from 256.7 m to 69.0 m, and the percentage of the geo-localized query pictures under a 50 m error requirement from 17.2% to 43.2% compared with the previous method. Another discovery of the proposed method is that, regarding the causes of reconstruction error, closer distances from the query cameras to the main objects in query pictures tend to produce smaller errors. The proposed method is not limited to small areas, and could be expanded to cities and larger areas owing to its flexible parameters.

Key words: street view images, geo-localization, image retrieval, 3D reconstruction, accuracy analysis

中图分类号: