MSRI : A Minimum Spatial Residual Iterative Algorithm for Vector Spatial Data Matching Based on Distance Decay

Résumé fourni par la source

ABSTRACT With the rapid advancement of geographic information technologies, multi‐source Vector Spatial Data have become increasingly accessible; however, achieving robust and automated matching remains a significant challenge due to structural inconsistencies and sensor misalignments. This study introduces the Minimum Spatial Residual Iterative (MSRI) algorithm, inspired by Tobler's Second Law of Geography, which integrates a spatial decay model with a residual‐based weighting scheme to guide iterative vector matching. Extensive experiments show that MSRI consistently outperforms classical methods such as Iterative Closest Point (ICP), Trimmed ICP (TrICP), and RANSAC + FPFH, particularly under irregular displacements, occlusions, and high data incompleteness. MSRI achieves an RMSE of 0.54 m, compared to 1.26–1.83 m for baselines, and improves F1‐scores by 12%–30%. Even with 70% data loss, MSRI maintains stable accuracy and reduces RMSE by up to 3.7× relative to TrICP, while ICP and RANSAC often fail to converge. A sensitivity analysis further confirms stable performance across practical parameter ranges, with optimal results obtained at α = 0.4 and β = 0.6. With its robustness, automation, and minimal preprocessing requirements, MSRI provides a practical solution for applications such as smart city construction, environmental monitoring, and large‐scale vector spatial data integration.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.