A System Model for Analyzing and Accumulating Construction Work Crew's Productivity Data Using Image Processing Technologies

Abstract

Recently, several studies pertaining to the measurement and analysis of construction productivity data have been attempted using image-processing technology. However, these studies have mainly focused on the recognition of individual labor and the materials at construction job sites. This research develops a new system model, which automatically analyzes and accumulates a construction work crew's productivity data using image processing technologies. It includes the three modules as follows: 1) the acquisition of a construction job site's video images and calibration for their recognition, 2) measurement and analysis of the construction work crew's productivity data, and 3) creation and utilization of a productivity database. New algorithms are proposed for each module. YCbCr settings are used to develop a calibration algorithm for improving the rate of recognition. Work sampling and video editing are used to develop an algorithm for measuring and analyzing the crew's productivity data. Then, an algorithm for productivity data accumulation and its utilization is proposed by utilizing the integration of the BIM Model. This developed system model is applied to a real construction site and validates its feasibility through two case studies.