Application of Computer Vision in the Construction Industry

Abstract

In recent years, there is a proliferation of computer vision applications in different industries. Computer vision enables machines to see visually and allows images or videos to be analysed with machine learning algorithms. An abundance of data can be generated daily from sensors, cameras, drones, autonomous robots, BIM (building information model) and other devices. With the help of AI, insights from big data can be drawn to improve construction design, workflow and schedule. In this paper, the focus is to look at computer vision applications that can be employed in the construction industry to resolve construction issues and improve productivity, security and safety of the worksites. Object detection using deep learning algorithm can be employed for inspection of construction material components. This may include contour detection, counting objects and determining size of objects. In terms of security on worksites, facial detection and recognition can be carried out at various points to ensure only qualified people are allowed to enter certain regions to carry out necessary work that include operating on a particular machine. Motion detection mechanism can be deployed in cameras to detect motion in certain regions for surveillance purposes. Real-time interactions of workers and machinery can be captured and workers’ emotions’ can be monitored to check on their wellbeing. Workers’ safety at worksite is always a top concern for companies. Drowsiness detection devices can be installed in cranes to monitor crane operators’ alertness at work. We will look at functionalities provided by OpenCV and discuss the setup, the use and boundary of applications such as object detection, motion detection, facial recognition and drowsiness detection that can be used in the construction industry.