BROWSE
Detail
Conference Paper
Voxel-Based YOLO Framework for Classifying Korean Traditional Wooden Architectural Components
Cited - time in 
Share 
Share
- Authors
- Minseo Kim, Nasrulloh Yahyozoda, Yongsik Lee, Seungjae Lee, Seungjoon Yang
- Issue Date
- 2025-11
- Citation
- 한국통신학회 종합 학술 발표회 (추계) 2025, pp.924-925
- Publisher
- 한국통신학회
- Language
- English
- Type
- Conference Paper
- Abstract
- Korean traditional wooden architecture, valued for its unique cultural significance, faces the risk of loss due to aging and limited restoration resources. This study proposes a YOLO-based 3D classification framework to categorize structural components from gongpo system using voxelized point clouds. By extending YOLO to 3D space, the proposed method enables accurate and automated classification of architectural elements, providing a scalable approach for preserving and revitalizing Korea’s wooden heritage.
- KSP Keywords
- 3D classification, 3D space, Automated classification, Cultural significance, Point clouds, Structural component, Voxel-based, Wooden architecture, architectural components