An assessment of the efficiency of spatial distances in linear object matching on multi-scale, multi-source maps

ABSTRACT

Distance in geospatial sciences has many applications, including the calculation of spatial similarity degree in object-matching problems. Various distances have so far been utilised for this purpose. However, no study has examined the efficiency of methods used for finding solutions for linear object matching in data sets with different or the same scales and sources. The present study investigated the efficiency of the most important and applicable spatial distances (13 types of distance methods) in vector data sets with different scales and sources. To this end, we employed three data sets of urban roads network of different sources with the scales of 1:2000, 1:5000 and 1:25,000. In the considered approach, the data sets are initially pre-processed to unify the format and coordinate system as well as removing topological errors. The corresponding objects in the data sets are then identified, and one-to-null, null-to-one, one-to-one, one-to-many, many-to-one and many-to-many relations are extracted. Ultimately, the method with the minimum dispersion in calculation of the distances between corresponding objects is selected as the efficient method. The results indicated that the short-line median and mean Hausdorff methods achieved higher efficiencies compared to the other employed methods. In addition to achieving a smaller variance compared to other introduced methods, these two methods are well capable of identifying one-to-many (many-to-one) and many-to-many relations.

KEYWORDS:

• Spatial distances
• linear object matching
• multi-scale
• multi-source data sets

Disclosure statement

No potential conflict of interest was reported by the author(s).