•  
  •  
 

Author Country (or Countries)

S. Korea

Abstract

Measuring productivity on a construction job site is not an easy task because collecting its reliable data consistently from the job site requires a lot of personnel efforts causing extra time and cost. In order to measure the job site productivity, for instance, basically two types of data such as man-hours and installed work quantities are required as an input and an output factor to calculate the values of the productivity. This paper suggests an efficient automatic man-hours measurement system for analyzing and collecting the data relating to the input factor by the analysis of video images instead of the direct observation of construction job sites by a work manager. The proposed method utilizes the image processing technologies for analyzing the video images of the construction related works operated on the job sites. An image processing based algorithm is developed for tracking the three different groups of construction work crews under the considerations of complex construction work environment. It includes the following main algorithms: BGS (background subtraction), worker detection, and worker tracking algorithm. The proposed method has been applied and verified under various indoor and outdoor experimental test environments. Around 84% of recognition rate for counting man-hours of construction work crews is achieved by the proposed algorithm. The data acquired by the proposed method can be used as an essential and valuable input data to analyze and control the productivity of construction job site subsequently.

Suggested Reviewers

N/A

Digital Object Identifier (DOI)

http://dx.doi.org/10.12785/amis/080343

Share

COinS