Computer Engineering and Applications ›› 2015, Vol. 51 ›› Issue (12): 138-143.

Previous Articles     Next Articles

Visual localization method for smart phones based on photogrammetry

ZOU Xiaodong1,2, XU Ying2, YUAN Hong2, TANG Yangyang2   

  1. 1.University of Chinese Academy of Sciences, Beijing 100049, China
    2.Academy of Opto-electronics, Chinese Academy of Sciences, Beijing 100094, China
  • Online:2015-06-15 Published:2015-06-30


邹晓东1,2,徐  颖2,袁  洪2,唐阳阳2   

  1. 1.中国科学院大学,北京 100049
    2.中国科学院 光电研究院,北京 100094

Abstract: Smart phones spread rapidly that localization methods using smart phones are becoming a concern. Smart phones always have a camera as well as an operational capability. So this paper presents a visual localization method based on photogrammetry. The method builds a database of reference images for service region in advance. When a phone enters this region, it shoots a real-time image using its camera, then two reference images which have same overlaps with the real-time image are searched by matching algorithm to get their overlapped points. According to the information of these two reference images, it can calculate the corresponding space-coordinates of these points using the fore-intersection. And then a resection method is used with these points to resolve location and attitude of the phone user. Results show that localization using this method can get an accuracy of just meters.

Key words: smart phone, visual localization, image match, photogrammetry

摘要: 随着智能手机的快速普及和广泛应用,利用手机进行定位的技术越来越受到人们关注。智能手机一般都具有摄像功能,并且有独立运算的能力,基于此,提出了一种基于摄影测量的手机视觉定位方法。该方法事先在提供视觉定位服务的区域拍摄基准图像,建立基准图像库;当手机用户进入定位服务区时,利用手机摄像头拍摄实时图像,在基准库中通过图像匹配方法搜索两幅与该图像具有相同重叠区域的基准图像,并得到三度重叠特征点。对这两幅基准图像,利用前方交会方法求出重叠区域特征点的空间坐标,再根据这些空间点坐标对实时图像进行后方交会,解算用户的位置和姿态。实验结果表明,该方法能够达到米级的定位精度。

关键词: 智能手机, 视觉定位, 图像匹配, 摄影测量