M:N Object matching on multiscale datasets based on MBR combinatorial optimization algorithm and spatial district

Abstract

Object matching is critical for updating, maintaining, integrating, and quality assessing spatial data. However, matching data are often obtained from different sources and have problems of positional discrepancy and different levels of detail. To resolve these problems, this article presents a multiscale polygonal object-matching approach, called the minimum bounding rectangle combinatorial optimization (MBRCO) with spatial district (SD). This method starts with the MBRCO algorithm and its enhancement using the SD to find corresponding MBRs of one-to-one, one-to-many, and many-to-many matching pairs. Then, it aligns the MBRs of the matching pairs to identify objectmatching pairs, which are evaluated using a matching criterion to find geometrically corresponding objects. Our approach was experimentally validated using two topographical datasets at 1:2k and 1:10k. The proposed approach outperforms the common two-way area overlap method and another method based on the contextual information and relaxation labeling algorithm. The proposed method achieved accurate aggregation of the many-tomany matching pairs under the positional discrepancy scenario.

A building polygonal object matching method based on minimum bounding rectangle combinatorial optimization and relaxation labeling

“…Nevertheless, this method cannot satisfactorily solve the problem of positional discrepancy. Liu et al (2018) proposed…”

“…• Divide the spatial districts (SDs): SDs are divided according to the matched pairs obtained in the first matching, which follows the work of Liu et al (2018).…”

“…where A M(a i ) is a function that translates polygon A a i according to the vector (Δx, Δy); A M(a i ) ∩ b j represents the overlapping area of A M(a i ) and A b j ; A γ is the threshold value and is often set as A γ = 0.3 (Fan et al, 2014;Liu et al, 2018;Rutzinger et al, 2009;Xu et al, 2017)…”

“…In addition, we also use threshold A ε in Algorithm 1, which is used to judge the similarity of MBRs of potential matching pairs. Liu et al (2018)…”

“…Currently, OSM is considered the most successful and popular VGI project, which has approximately 1.75 million contributors and shares more than 512 million building footprint data (by 2021‐Q4). The MBRCO algorithm was proposed by Liu et al (2018) and has the following characteristics. First, the MBRCO algorithm detects a matching pair based on the similarity of their shapes.…”