Multi-Range Conditional Random Field for Classifying Railway Electrification System Objects Using Mobile Laser Scanning Data

Jung, Jaewook; Chen, Leihan; Sohn, Gunho; Luo, Chao; Won, Jong-Un

doi:10.3390/rs8121008

Cited by 18 publications

(17 citation statements)

References 27 publications

(53 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The regular supervised machine learning approach has been introduced to recognize railway objects. For instance, Jung et al recognized ten categories of railway electrification system objects from the MLS point cloud by using a multi-range CRF model [7] or a multi-scale hierarchical CRF model [8]. However, few studies introduced the deep learning approach to recognized railway objects from MLS data.…”

Section: Related Workmentioning

confidence: 99%

See 1 more Smart Citation

A Deep Learning-Based Method for Overhead Contact System Component Recognition Using Mobile 2D LiDAR

Chen

Lin

et al. 2020

Sensors

View full text Add to dashboard Cite

The overhead contact system (OCS) is a critical railway infrastructure for train power supply. Periodic inspections, aiming at acquiring the operational condition of the OCS and detecting problems, are necessary to guarantee the safety of railway operations. One of the OCS inspection means is to analyze data of point clouds collected by mobile 2D LiDAR. Recognizing OCS components from the collected point clouds is a critical task of the data analysis. However, the complex composition of OCS makes the task difficult. To solve the problem of recognizing multiple OCS components, we propose a new deep learning-based method to conduct semantic segmentation on the point cloud collected by mobile 2D LiDAR. Both online data processing and batch data processing are supported because our method is designed to classify points into meaningful categories of objects scan line by scan line. Local features are important for the success of point cloud semantic segmentation. Thus, we design an iterative point partitioning algorithm and a module named as Spatial Fusion Network, which are two critical components of our method for multi-scale local feature extraction. We evaluate our method on point clouds where sixteen categories of common OCS components have been manually labeled. Experimental results show that our method is effective in multiple object recognition since mean Intersection-over-Unions (mIoUs) of online data processing and batch data processing are, respectively, 96.12% and 97.17%.well record geometric details of surrounding objects [5]. Therefore, it is appropriate to apply MLS systems to OCS inspections. The operational condition of OCS can be acquired by analyzing the MLS point cloud instead of manual measurement. Recognizing the point cloud of OCS as a critical task of the data analysis has been studied in previous studies (e.g., [6][7][8]). Mobile 2D LiDAR is a special kind of MLS system applied to railway inspections. Figure 1 shows an instance of the mobile 2D LiDAR used to scan the OCS infrastructure. As for this kind of MLS system, a point cloud is built up by integrating points at each 2D scan line. Fully understanding the point cloud is significant for automatic inspections and intelligent diagnoses. Thus, this study focuses on multiple OCS component recognition with mobile 2D LiDAR. However, it is a difficult task because the similarities and the various associations among OCS components make scenes complex [8]. In this case, model-driven and data-driven methods become incompetent because rules and features for recognizing specific objects are difficult to be designed by human beings. Fortunately, the success of deep learning-based image recognition (e.g., [9][10][11]) has promoted the fast development of deep learning-based point cloud segmentation, which provides novel means to understand point cloud with semantics.

show abstract

Section: Related Workmentioning

confidence: 99%

“…The operational condition of OCS can be acquired by analyzing the MLS point cloud instead of manual measurement. Recognizing the point cloud of OCS as a critical task of the data analysis has been studied in previous studies (e.g., [6][7][8]). Mobile 2D LiDAR is a special kind of MLS system applied to railway inspections.…”

mentioning

confidence: 99%

A Deep Learning-Based Method for Overhead Contact System Component Recognition Using Mobile 2D LiDAR

Chen

Lin

et al. 2020

Sensors

View full text Add to dashboard Cite

show abstract

“…Sawadisavi et al [21] used the machine vision method to inspect rails using images. Jung et al [37] used a supervised learning method to classify the Mobile Laser Scanning (MLS) data into ten target classes representing overhead wires, movable brackets, and poles. Image based methods cannot generate a high-precision railroad map but produce images with unique advantages in distortion detection.…”

Section: Related Workmentioning

confidence: 99%

A Fast Algorithm for Rail Extraction Using Mobile Laser Scanning Data

et al. 2018

View full text Add to dashboard Cite

Railroads companies conduct regular inspections of their tracks to maintain and update the geographic data for railway management. Traditional railroad inspection methods, such as onsite inspections and semi-automated analysis of imagery and video data, are time consuming and ineffective. This study presents an automated effective method to detect tracks on the basis of their physical shape, geometrical properties, and reflection intensity feature. This study aims to investigate the feasibility of fast extraction of railroad using onboard Velodyne puck data collected by mobile laser scanning (MLS) system. Results show that the proposed method can be executed rapidly on an i5 computer with at least 10 Hz. The MLS system used in this study comprises a Velodyne puck/onboard GNSS receiver/inertial measurement unit. The range accuracy of Velodyne puck equipment is 2 cm, which fulfills the need of precise mapping. Notably, positioning STD is lower than 4 cm in most areas. Experiments are also undertaken to evaluate the timing of the proposed method. Experimental results indicate that the proposed method can extract 3D tracks in real-time and correctly recognize pairs of tracks. Accuracy, precision, and sensitivity of total test area are 99.68%, 97.55%, and 66.55%, respectively. Results suggest that in a multi-track area, close collaboration between MLS platforms mounted on several trains is required. terrestrial laser scanning (TLS), and MLS is common in recent years because laser scanners provide massive high-quality range measurements with intensity features in a fast manner. Many remarkable practices are made in MLS, ALS, and TLS. The ALS technology has been commercialized in the generation of high-precision digital elevation and surface models [3]. The first MMS [4] used onboard satellite receivers and stereo cameras to capture road images. At that time, precise point positioning [5] technology and differential positioning [6] technology were not introduced yet. IMUs are expensive and the application of mobile mapping systems (MMSs) is limited. Currently, Crucial high-end MMSs such as the Optech Lynx mobile mapping system, Trimble MX-9 MLS system, and RIGEL VMX-2HA MLS system could capture over millions of points every second. Commercial MMSs are applied to point cloud data acquisition of urban road networks and railroad corridors. However, commercial MMSs do not provide a user programming interface, which makes object detection and extraction applications based on them impossible. Consequently, previous studies were focused on the combined use of those devices [7].In general terms, an MLS platform is a vehicle-mounted MMS that is integrated with multiple sensors (e.g., GNSS antenna, IMU, digital cameras) and equipped with a centralized computing system for data synchronization and management. Existing urban reconstruction algorithms using data from different source (e.g., ALS, TLS, MLS) obtain convincing results in urban object modeling and management. Data collected by vehicle-mounted MMSs can be used to detect o...

show abstract

“…In order to mitigate the computational burden encountered in optimization, a necessary clustering like super pixel grouping (Fulkerson et al, 2009), or line extraction (Jung et al, 2016) is drawn on the data for large scale classification applications.…”

Section: Introductionmentioning

confidence: 99%

Semantic Segmentation of Indoor Point Clouds Using Convolutional Neural Network

Babacan

Chen

Sohn

2017

ISPRS Ann. Photogramm. Remote Sens. Spatial Inf. Sci.

View full text Add to dashboard Cite

ABSTRACT:As Building Information Modelling (BIM) thrives, geometry becomes no longer sufficient; an ever increasing variety of semantic information is needed to express an indoor model adequately. On the other hand, for the existing buildings, automatically generating semantically enriched BIM from point cloud data is in its infancy. The previous research to enhance the semantic content rely on frameworks in which some specific rules and/or features that are hand coded by specialists. These methods immanently lack generalization and easily break in different circumstances. On this account, a generalized framework is urgently needed to automatically and accurately generate semantic information. Therefore we propose to employ deep learning techniques for the semantic segmentation of point clouds into meaningful parts. More specifically, we build a volumetric data representation in order to efficiently generate the high number of training samples needed to initiate a convolutional neural network architecture. The feedforward propagation is used in such a way to perform the classification in voxel level for achieving semantic segmentation. The method is tested both for a mobile laser scanner point cloud, and a larger scale synthetically generated data. We also demonstrate a case study, in which our method can be effectively used to leverage the extraction of planar surfaces in challenging cluttered indoor environments.

show abstract

Multi-Range Conditional Random Field for Classifying Railway Electrification System Objects Using Mobile Laser Scanning Data

Cited by 18 publications

References 27 publications

A Deep Learning-Based Method for Overhead Contact System Component Recognition Using Mobile 2D LiDAR

A Deep Learning-Based Method for Overhead Contact System Component Recognition Using Mobile 2D LiDAR

A Fast Algorithm for Rail Extraction Using Mobile Laser Scanning Data

Semantic Segmentation of Indoor Point Clouds Using Convolutional Neural Network

Contact Info

Product

Resources

About