Tool-Based Automation in 3D Point Cloud Processing – A Review

Abstract

This paper provides a systematic review of tool-based automation in 3D point cloud processing, focusing on open-source platforms and their capabilities to facilitate spatial workflows. The irregular nature of point clouds and their increasing volume are increasingly utilized in geomatics, construction, agriculture, and heritage documentation, making manual processing impractical. Scalability, accuracy, and usability, therefore, require automation. It was a systematic review based on Scopus, IEEE Xplore, and Web of Science (2019-2025). A total of 36 articles were located using keywords, including automation, machine learning, open-source tools, and point cloud processing, which applied automation in a tool or platform. The results are organized into three domains: segmentation, classification, and reconstruction. In segmentation, voxel-based partitioning and transformer networks have been implemented in platforms such as CloudCompare and Open3D to enhance scalability and detail capture. In classification, tools are increasingly instruments that integrate machine learning with other forms of contextual reasoning, and forestry and UAV applications are examples of their potential. Reconstruction studies show an increasing relationship between BIM and heritage workflows, enabled by tools such as Cloud2BIM and Open3DGen. Although these methods reduce manual labor and increase efficiency, they still have limitations in terms of compatibility, benchmark standardization, and cross-domain applicability. The review, in general, synthesizes progress but also points out a lack of consistency in current advances. It also helps guide future point cloud research by clarifying the strengths and gaps in creating more compatible, benchmarked, and domain-flexible tool-based automation.