A systematic scheme to maintain multiple characteristics for effective polygon rasterization
Chen Zhou Manchun Li Department of Geographic Information Science,School of Geography and Ocean Science,Nanjing University,Nanjing,Jiangsu Province,P. R. ChinaChen Zhou is a researcher at Department of Geographic Information Science,Nanjing University. His research topics include geospatial analysis and high-performance geocomputation. He devised the main conceptual ideas,performed the experiments,and wrote the manuscript.Manchun Li is a professor at Department of Geographic Information Science,Nanjing University. His research interests include cartography and geographical information science,and parallel computing. He worked on the technical details,and provided sufficient suggestions for revisions.
DOI: https://doi.org/10.1080/13658816.2024.2391982
2024-08-20
International Journal of Geographical Information Science
Abstract:Polygon characteristics, including area, shape, and topology, may be lost during rasterization, leading to inaccurate analyses. Maintaining multiple characteristics remains a challenging multi-objective optimization problem. This study introduces improved strategies for flexibly maintaining multiple characteristics (ISMMC). Errors related to area (A), shape (S), and topology (T) are addressed using six prioritized sequences: AST, ATS, TAS, TSA, STA, and SAT, where the order of the letters denotes the sequence of error correction. The importance (I) of land-use types (LUT) is also considered when developing a strategy based on the TSA (ITSA) that maintains characteristics for important LUTs. The ISMMC achieved a total improved accuracy (IA) of ≥0.05 for 98.67% of 150 tested cases. The STA, SAT, TAS, and TSA yielded the highest area, shape, topological, and total accuracies. The STA achieved a summed AI increase ≥20.36% compared to the existing version. The runtime was ≥85.46 times the scanline runtime, and the threshold cell size (≤50 m) balanced the IA and runtime. The ITSA improved accuracy by 83.33% and 6.67% of solutions when preferentially maintaining one or two important LUTs, outperforming the existing STA. Overall, these developments improve accuracy with flexibility in rasterization results and subsequent raster applications.
geography, physical,computer science, information systems,information science & library science