Confidential — Stefan Michaelcheck Only

A Method for Recognizing I-Shaped Building Patterns Utilizing Multi-Scale Data and Knowledge Graph

2026application demonstrationapplicationmethod

Shenglu Xu, Bo Qiang, Pengpeng Li, Tao Liu, Ping Du, Wenning Wang

ISPRS International Journal of Geo-Information

https://doi.org/10.3390/ijgi15010023OpenAlex: W7118194615
10
URLs Found
0
Internal Citations
6
Authors
usable
Abstract Quality
GPT-5.5 Abstract Analysis

Problems Identified (2)

Low-robustness geometric building pattern recognition: Traditional isolated-object geometric feature extraction methods struggle to capture complex visually significant urban building patterns and suffer from low accuracy and robustness.

Low-robustness geometric building pattern recognition: Traditional isolated-object geometric feature extraction methods struggle to capture complex visually significant urban building patterns and suffer from low accuracy and robustness.

Proposed Solutions (2)

Multi-scale knowledge-graph pattern recognition: The paper proposes recognizing I-shaped building patterns by extracting multi-scale inter-building relationships from SBR-based geometric and spatial rules and encoding them in a knowledge graph for rule-based reasoning.

Multi-scale knowledge-graph pattern recognition: The paper proposes recognizing I-shaped building patterns by extracting multi-scale inter-building relationships from SBR-based geometric and spatial rules and encoding them in a knowledge graph for rule-based reasoning.

Results (3)

Improved I-shaped pattern recall:

Human-visual-principle alignment:

Improved I-shaped pattern recall:

Research Domain

Urban design and spatial analysis / geospatial building pattern recognition

← Back to all papers