EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

G3H-GraphSim: A Reproducible 3D GeoHash Cell-Graph Workflow for Semantic LiDAR Analysis

Kyusoo Chong

Remote Sensing · 17.11.2026 · 18 (22) · 3939

DOI: 10.3390/rs18223939 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Remote Sensing, online erschienen am 17.11.2026. Verfasst von Kyusoo Chong. Die Arbeit nennt 23 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: KICT Research Program.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Reliable infrastructure graph analysis requires transparent links between point-cloud observations, semantic annotations, spatial addresses, and graph decisions. This study evaluates G3H-GraphSim, a reproducible three-dimensional GeoHash cell-graph workflow on five adjacent Gwangju LAS–annotation tiles, using shared-origin metric grids, stage-level computational measurements, and tile-resolved paired routing statistics. Tile-local level-7 cells have unequal physical dimensions and largest-component shares of 26.0–95.1%. In a pooled grid with 4 m × 4 m horizontal cells, six-neighbor largest-component shares are 24.7% at 0.5 m and 26.5% at 1.0 m vertical resolution, showing that common addressing alone does not ensure graph continuity. A noncontemporaneous OpenStreetMap cross-check places 95.3% of retained road-cell centers within 10 m of mapped centerlines, although only 63.5% of selected reference-line length is covered. On the same 150 start–goal pairs, integrated routing reduces pooled mean speed-bump proximity and high-LiDAR-density cell entries by 18.0% and 28.5%, respectively, at a 0.6% increase in mean route length, with appreciable tile-to-tile variation. Stage-level tests show lower amortized batch exact-key lookup time and smaller key-array payload for packed integers than for padded Base32 strings. The results demonstrate traceable sensitivity to spatial resolution, representation, and routing policy, without establishing field-calibrated safety or deployed-system performance.

Förderung

  • KICT Research Program

Zitieren

APA

Chong, K. (2026). G3H-GraphSim: A Reproducible 3D GeoHash Cell-Graph Workflow for Semantic LiDAR Analysis. Remote Sensing, 18(22), 3939. https://doi.org/10.3390/rs18223939

Vancouver

Chong K. G3H-GraphSim: A Reproducible 3D GeoHash Cell-Graph Workflow for Semantic LiDAR Analysis. Remote Sensing. 2026;18(22):3939. doi:10.3390/rs18223939

BibTeX
@article{Chong2026,
  author = {Chong, Kyusoo},
  title = {G3H-GraphSim: A Reproducible 3D GeoHash Cell-Graph Workflow for Semantic LiDAR Analysis},
  journal = {Remote Sensing},
  year = {2026},
  volume = {18},
  number = {22},
  pages = {3939},
  doi = {10.3390/rs18223939}
}
Kostenlos starten →