Semantic Segmentation of Point Clouds With Pointnet and Kpconv Architectures Applied to Railway Tunnels

Abstract

Abstract. Transport infrastructure monitoring has lately attracted increasing attention due to the rise in extreme natural hazards posed by climate change. Mobile Mapping Systems gather information regarding the state of the assets, which allows for more efficient decision-making. These systems provide information in the form of three-dimensional point clouds. Point cloud analysis through deep learning has emerged as a focal research area due to its wide application in areas such as autonomous driving. This paper aims to apply the pioneering PointNet, and the current state-of-the-art KPConv architectures to perform scene segmentation of railway tunnels, in order to validate their employability over heuristic classification methods. The approach is to perform a multi-class classification that classifies the most relevant components of tunnels: ground, lining, wiring and rails. Both architectures are trained from scratch with heuristically classified point clouds of two different railway tunnels. Results show that, while both architectures are suitable for the proposed classification task, KPConv outperforms PointNet with F1-scores over 97% for ground, lining and wiring classes, and over 90% for rails. In addition, KPConv is tested using transfer learning, which gives F1-scores slightly lower than for the model training from scratch but shows better generalization capabilities.

Railway Lidar Semantic Segmentation With Axially Symmetrical Convolutional Learning

“…Exploring deep-learning approaches, Soilán et al (2020) apply PointNet (Qi et al, 2017a) and KPCNN (Thomas et al, 2019) to segmentation of railway tunnels point cloud data. Although the test environment is simple, their work shows hopeful results for full infrastructure segmentation using deep-learning-based approaches.…”

Enhancing Railway Detection by Priming Neural Networks with Project Exaptations

“…An existing model was adapted and compared with a newly trained ML model. Previous results from Sánchez-Rodríguez et al [18] were used as training and test data [26]. The transfer learning is homogenous, and the task is comparable, as only a so-called class mismatch occurs [24].…”

Automatic Point Cloud Semantic Segmentation of Complex Railway Environments

“…2021, 13, x FOR PEER REVIEW 3 of 41 it is not applicable to more complex and variable scenarios [21]. In other previous studies, [22] a different method is introduced with the same objectives, using Pointnet [23] and KPConv [24], but applicable only to tunnels. In that vein, a methodology is developed that is applicable to complex areas and tested at 90 km but only for the delineation of railway lanes in order to generate IFC alignment models [25].…”

“…In previous research, an automatic detection method is presented, along with the decomposition of a railway from cloud points in tunnels, with good results using support vector machine algorithms (SVMs), but it is not applicable to more complex and variable scenarios [21]. In other previous studies, [22] a different method is introduced with the same objectives, using Pointnet [23] and KPConv [24], but applicable only to tunnels. In that vein, a methodology is developed that is applicable to complex areas and tested at 90 km but only for the delineation of railway lanes in order to generate IFC alignment models [25].…”