Computer Vision Techniques in Construction: A Critical Review
- PDF / 1,276,637 Bytes
- 15 Pages / 595.276 x 790.866 pts Page_size
- 22 Downloads / 374 Views
ORIGINAL PAPER
Computer Vision Techniques in Construction: A Critical Review Shuyuan Xu1 · Jun Wang2 · Wenchi Shou3 · Tuan Ngo4 · Abdul‑Manan Sadick2 · Xiangyu Wang5,6 Received: 9 March 2020 / Accepted: 30 September 2020 © CIMNE, Barcelona, Spain 2020
Abstract Computer vision has been gaining interest in a wide range of research areas in recent years, from medical to industrial robotics. The architecture, engineering and construction and facility management sector ranks as one of the most intensive fields where vision-based systems/methods are used to facilitate decision making processes during the construction phase. Construction sites make efficient monitoring extremely tedious and difficult due to clutter and disorder. Extensive research has been carried out to investigate the potential to utilise computer vision for assisting on-site managerial tasks. This paper reviews studies on computer vision in the past decade, with a focus on state-of-the-art methods in a typical vision-based scheme, and discusses challenges associated with their application. This research aims to guide practitioners to successfully find suitable approaches for a particular project.
1 Introduction Computer vision as an interdisciplinary scientific field involves using computers to gain a detailed understanding of visual data, which is a similar approach to that of human visual systems. This field falls into the broad category of artificial intelligence and its applications have tackled a wide range of problems. Lowe [1] lists several industrial applications of computer vision including automotive driver assistance and traffic management, eye and head tracking for consumer research, real-time vision for surgical applications, and industrial automation and inspection in semiconductor * Jun Wang [email protected] * Xiangyu Wang [email protected] 1
School of Design and the Built Environment, Curtin University, Bentley, Australia
2
School of Architecture and Built Environment, Deakin University, Geelong, Australia
3
School of Computing, Engineering and Mathematics, Western Sydney University, Parramatta, Australia
4
Department of Infrastructure Engineering, The University of Melbourne, Parkville, Australia
5
School of Civil Engineering and Architecture, East China Jiaotong University, Nanchang, China
6
Australasian Joint Research Centre for Building Information Modelling, Curtin University, Bentley, Australia
manufacturing. Forsyth and Ponce [2] categorized these processes into three sequential stages, which illustrate the extraction of basic features in visual data (e.g., colour, texture and depth information), processing, and semantic information generation (as shown in Fig. 1). Computer vision is linked to the architecture, engineering and construction and facility management (AEC/FM) sector and the last decade has witnessed a growing trend of its applications on construction sites. Due to their intrinsic nature of being cluttered and unordered, jobsites are prone to management failures such as unsati
Data Loading...
