Deep Unsupervised Segmentation of Log Point Clouds

2025-03-18

AI Summary

To address the high cost of X-ray CT systems in log measurement at sawmills, this paper proposes an unsupervised surface segmentation method for log point clouds to enable internal structure inference. The method builds upon the Point Transformer architecture and introduces a novel unsupervised loss function that jointly incorporates cylindrical geometric priors and models wood-shape variability—eliminating the need for labeled data while achieving accurate surface-point identification. Evaluated on real log datasets, it significantly improves surface-point classification accuracy and demonstrates strong generalization to other cylindrical objects. This work establishes a new paradigm for low-cost, rapid 3D log measurement and represents the first study to integrate Point Transformers with explicit geometric constraints for unsupervised segmentation of log point clouds.

Technology Category

• Machine Learning: Calibration & Uncertainty Quantification
• Computer Vision: Segmentation
• Search and Optimization: Mixed Discrete/Continuous Search

Application Category

• Semantics and Knowledge: Methods, algorithms and applications for the development of semantic models, knowledge graphs and other forms of structured data models with machine-interpretable semantics
• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications
• Security and Privacy: Large-scale security measurements

Abstract

In sawmills, it is essential to accurately measure the raw material, i.e. wooden logs, to optimise the sawing process. Earlier studies have shown that accurate predictions of the inner structure of the logs can be obtained using just surface point clouds produced by a laser scanner. This provides a cost-efficient and fast alternative to the X-ray CT-based measurement devices. The essential steps in analysing log point clouds is segmentation, as it forms the basis for finding the fine surface details that provide the cues about the inner structure of the log. We propose a novel Point Transformer-based point cloud segmentation technique that learns to find the points belonging to the log surface in unsupervised manner. This is obtained using a loss function that utilises the geometrical properties of a cylinder while taking into account the shape variation common in timber logs. We demonstrate the accuracy of the method on wooden logs, but the approach could be utilised also on other cylindrical objects.

Problem

• Accurate measurement of wooden logs in sawmills
• Unsupervised segmentation of log surface point clouds
• Optimizing sawing process using laser scanner data

Innovation

• Point Transformer-based unsupervised segmentation technique
• Utilizes cylinder geometry for log surface detection
• Cost-efficient alternative to X-ray CT measurements