P2-3-1 
MOTION ANALYSIS ON THE CONSTRUCTION PLANT USING SEQUENTIAL IMAGES 
Sosuke YOSMDA, Hirofumi CHDCATSU 
Department of Civil Engineering 
Tokyo Denki University 
JAPAN 
sosuke@g. dendai ac.jp 
Commission V, Working Group IC V/III 
KEY WORDS: Ergoma System, Morphology, Feature Extraction, Work Analysis, Motion Analysis, Sequential Images 
ABSTRACT 
Recently, reduction of the construction cost and insufficiency of skilled hands are becoming serious problems on the construction plant 
from the point of view of the work efficiency. With this motive, the authors have been developing an ergoma system for improvement of 
working environment or evaluation of the work efficiency on the construction plant using sequential images. However, an extraction of 
the feature points under insufficient image quality that was caused by radiation of welding or burnishing is still issue that needs to be 
resolved. 
This paper describes an automated extraction method of the feature points using opening in morphological image processing, and shows 
human motion analysis on the construction plant. 
1 INTRODUCTION 
Recently, reduction of the construction cost and insufficiency of 
skilled hands are becoming serious problems on the construction 
plant from the point of view of the work efficiency. In these 
circumstances, an ergoma system has recently received more 
attention from the point of view of the improvement of working 
environment or evaluation of the work efficiency. An ergoma is a 
coined word by ergonomics and management. In order to develop 
an ergoma system, a video camera is expected to become useful 
tool since a video image sequences often give important 
information to human motion analysis. However, generally 
welding and burnishing on the construction plant are attached 
with strong radiation, and radiation causes insufficient image 
quality. Then, image processing procedure for extracting the 
human feature points during welding or burnishing should be 
resolved. For this issue, welding and burnishing were performed 
on the construction plant and these working sequential images 
were taken using two video cameras. 
This paper describes an automated extraction method of the 
feature points using opening in morphological image processing, 
and shows human motion analysis during welding and burnishing 
2 EXPERIMENTS 
In order to develop an ergoma system on the construction plant, 
welding and burnishing for horizontal and vertical steel pipes 
were performed respectively and sequential images for each work 
were taken using two video cameras. Fluorescence tapes were put 
on the head (helmet), waist and arms as feature points. However, 
fluorescence tape was put on the head (helmet) are used for 
developing an automatic extraction method in this paper. Figure 
1 shows the test site. Figure 2 shows an image acquired by video 
camera, and the square targets in the Figure 2 are control points 
for camera calibration. Figure 3 shows the system configuration 
that was used in this investigation. 
In this system, the LVR (Laser Videodisc Recorder) has function 
for playing back and stop per a frame and connecting with 
another machines and tools such as PC or Video camera. Then, 
the original video images copied to the digital CRVdisc 
(Component Recording Videodisc) for the LVR (Laser Videodisc 
Horizontal steel pipe 
Vertical steel pipe 
Left video camera 
Right video camera 
Figure 1 The test site. 
Figure 2 Image on the test site