Progressive merging of blocks taking into account functional semantics and grid features

Yue Chen,Haizhong Qian,Xiao Wang
DOI: https://doi.org/10.1080/15230406.2024.2353767
IF: 2.354
2024-05-30
Cartography and Geographic Information Science
Abstract:Continuous-scale representation of urban blocks is critical for enriching the web-based mapping experience in the modern era. Traditionally, continuous-scale merging methods for surface elements have been employed to deal with landscape features, such as patches and land-cover data. In comparison, the artificially constructed urban blocks offer complex semantics and unique shapes. Traditional methods face challenges in maintaining their fundamental characteristics, which can result in improper merging outcomes. Therefore, this paper proposes a progressive merging method that takes into account the functional semantics and grid features of blocks. Initially, the POI data is used to obtain the functional semantic vectors of the blocks for creating basic information about the starting map. The blocks are then abstracted into a graphical structure whereby a breadth-first algorithm is employed to generate their groups; thereby, forming the target map. The map-based geometric and semantic merging cost formula has been devised to calculate the optimal merging sequence by the A* algorithm. Experiments have been performed by defining the main urban area of Beijing city as the study area. The comparative analysis indicates that the proposed method can obtain more reasonable progressive merging results.
geography
What problem does this paper attempt to address?