Status quo and open challenges in vision-based sensing and tracking of temporary resources on infrastructure construction sites

被引:142
作者
Teizer, Jochen [1 ]
机构
[1] RAPIDS Construct Safety & Technol Lab, Munich, Germany
基金
美国国家科学基金会;
关键词
Building information modeling; Computer vision and machine learning; Resource location tracking and progress monitoring; Safety and health; Sensors: photo and video cameras; unmanned aerial vehicles; Surveying: laser scanning photo- and videogrammetry; VISUALIZATION TECHNOLOGY; EQUIPMENT; FUSION; SAFETY; RECONSTRUCTION; SEGMENTATION; RECOGNITION; OPERATIONS; RETRIEVAL; PROGRESS;
D O I
10.1016/j.aei.2015.03.006
中图分类号
TP18 [人工智能理论];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
Modern construction projects require sufficient planning and management of resources to become successful. Core issues are tasks that deal with maintaining the schedule, such as procuring materials, guaranteeing the supply chain, controlling the work status, and monitoring safety and quality. Timely feedback of project status aids project management by providing accurate percentages of task completions and appropriately allocating resources (workforce, equipment, material) to coordinate the next work packages. However, current methods for measuring project status or progress, especially on large infrastructure projects, are mostly based on manual assessments. Recent academic research and commercial development has focused on semi- or fully-automated approaches to collect and process images of evolving worksites. Preliminary results are promising and show capturing, analyzing, and documenting construction progress and linking to information models is possible. This article presents first an overview to vision-based sensing technology available for temporary resource tracking at infrastructure construction sites. Second, it provides the status quo of research applications by highlighting exemplary case. Third, a discussion follows on existing advantages and current limitations of vision based sensing and tracking. Open challenges that need to be addressed in future research efforts conclude this paper. (C) 2015 Elsevier Ltd. All rights reserved.
引用
收藏
页码:225 / 238
页数:14
相关论文
共 99 条
[1]   Formalization and automation of time-space conflict analysis [J].
Akinci, B ;
Fischen, M ;
Levitt, R ;
Carlson, R .
JOURNAL OF COMPUTING IN CIVIL ENGINEERING, 2002, 16 (02) :124-134
[2]  
[Anonymous], P CONSTR RES C 2010, DOI DOI 10.1061/41109(373)2
[3]  
[Anonymous], 2000, Multiple View Geometry in Computer Vision
[4]  
[Anonymous], 2014, COMPREHENSIVE METHOD
[5]  
[Anonymous], TECHNOLOGY DESIGN PR
[6]  
[Anonymous], 2009, PERFORMANCE EVALUATI
[7]   Non-Rigid Object Localization and Segmentation Using Eigenspace Representation [J].
Arif, Omar ;
Vela, Patricio Antonio .
2009 IEEE 12TH INTERNATIONAL CONFERENCE ON COMPUTER VISION (ICCV), 2009, :803-808
[8]   A tutorial on particle filters for online nonlinear/non-Gaussian Bayesian tracking [J].
Arulampalam, MS ;
Maskell, S ;
Gordon, N ;
Clapp, T .
IEEE TRANSACTIONS ON SIGNAL PROCESSING, 2002, 50 (02) :174-188
[9]   Ensemble tracking [J].
Avidan, Shai .
IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE, 2007, 29 (02) :261-271
[10]   Similarity Metrics for Categorization: from Monolithic to Category Specific [J].
Babenko, Boris ;
Branson, Steve ;
Belongie, Serge .
2009 IEEE 12TH INTERNATIONAL CONFERENCE ON COMPUTER VISION (ICCV), 2009, :293-300