A REVIEW OF POINT CLOUD SEGMENTATION OF ARCHITECTURAL CULTURAL HERITAGE

Abstract

Abstract. With the continuous development of laser scanning technology, the model reconstruction and information management system of architectural cultural heritage are gradually taking shape. Among them, point cloud segmentation of architectural cultural heritage is important for Historic Building Information Model (HBIM), disease extraction and analysis, heritage restoration and other research. This paper mainly focuses on the systematic analysis and summary of the point cloud segmentation of architectural cultural heritage, introduces related concepts, and summarizes the segmentation methods from three aspects: traditional methods, machine learning based on artificial features, and deep learning. In addition, this paper summarizes the evaluation metrics and public datasets of semantic segmentation commonly used today, and further elucidates the relevant applications after segmentation. Finally, this paper analyzes and prospects the main problems and future development trends. This review aims to provide a useful reference for relevant researchers in the field of architectural cultural heritage protection.