A New Method for Automatic Large Scale Map Updating Using Mobile Mapping Imagery. (26th April 2013)

• Record Type: Journal Article
• Authors: Ou, Jianliang; Qiao, Gang; Bao, Feng; Wang, Weian; Di, Kaichang; Li, Rongxing
• Is Part Of: Photogrammetric record. Volume 28:Number 143(2013:Sep.)
• Page Start: 240
• Page End: 260
• Publication Date: 2013-04-26
• Subjects: building recognition -- land‐based mobile mapping -- large scale map updating -- line segment matching -- vanishing point -- vertical lines; Photogrammetry -- Periodicals; 526.982
• DOI: 10.1111/phor.12018
• Languages: English
• ISSNs: 0031-868X

Abstract

Abstract: Land‐based mobile mapping system (MMS) technology provides an innovative way to update geospatial information, characterised by rapid data collection and direct georeferencing. This paper presents a new method for automatically updating large scale maps using sequential MMS imagery. A major component of the technique is the extraction of vertical lines from such imagery through an improved Hough transform and vanishing point classification. These lines are stereomatched using an epipolar constraint and random sample consensus (RANSAC) strategy, with object space positions obtained by photogrammetric intersection. Buildings are recognised based on these vertical lines in conjunction with basic polygon templates, allowing maps to be updated.