Automatic Collection And Visualization Of Indoor Paving Progress Information Supported By BIM And Computer Vision

Progress control is recognized as one of the major goals of construction project management.Effective progress control relies heavily on timely and correct decision-making by project managers.For project managers,accurate,timely and intuitive information is the key to successful decision-making during progress control on construction site.However,most of construction projects still use the traditional method(e.g.manual inspection)to collect the actual project progress information from construction site.Manual inspection method has many drawbacks,such as work-intensive,time consuming,error prone and great reliance on the experience of workers.In addition,current progress reporting uses textual reports and progress curve,which has little help to the decision-making.Based on referred background,this paper presents a method that combines computer vision with building information modeling(BIM)for automated progress monitoring.This method can automatically and accurately measure the actual progress information of the construction site,and transmit the real-time progress information to the cloud in a visualized form.With the actual and real-time progress information,project managers can know the progress of construction site in time and make decisions easily.The proposed method includes five modules: training of image recognition classifier,detection of edge contour of target component,camera calibration,extraction information from BIM model and visual display of actual progress.Firstly,an image database is built with thousands of tile images,and a variety of LBP feature extraction methods and support vector machine(SVM)are used to train a tile classifier with good performance,of which accuracy reaches 91.17%.The improved edge detection algorithm detects the boundaries of completed tiles in given images.After that,the boundary line coordinates are converted from the image pixel coordinates to the real-world coordinate system through camera calibration.Then,using the information of camera location and room profile information extracted from BIM model,the actual tile area can be calculated automatically.Finally,the actual progress is highlighted in the room floor plan and the result is delivered to the BIM cloud simultaneously.The proposed method has been tested in a real indoor construction site.The experimental results indicates that this method can effectively realize the real-time automatic quantity calculation.In addition,this method can collect a large amount of progress information,which provides the possibility of accurate statistics and analysis on labor efficiency.