110 
Application of a Video Theodolite System for Sports Dynamics 
Hirofumi CHIKATSU Department of Civil Engineering, Tokyo Denki University 
Hatoyama, Saitama,350-03, JAPAN, FAX +82-492-96-6501 
Shunji MURAI Division of Computer Science, Asian Institute of Technology 
Bangkok,10501, THAILAND, FAX +66-2-516-2126 
KEY WORDS : Video Theodolite, Sequential Images, Camera Orientation, Sports Dynamics 
Abstract 
For the understanding of the dynamics of human motion from general TV or video image sequences, the camera 
orientation for each sequential image becomes an important subject. By using a video theodolite system as proposed in 
this paper, w and « for each sequential image can be acquired in real time as the vertical and horizontal angles, while ¢ 
is assumed to be 0 degrees since the theodolite is levelled. 
This paper describes the video theodolite system and its application for dynamic analysis of human motion with 
sequential images. Finally, based on experimental results, the validity of utilizing this systemin the dynamic analysis of 
human motion is indicated. 
1. Introduction 
Video imagery gives important information for dynamic analysis of human motion in the field of sports training or 
rehabilitation. For the understanding of the dynamics of human motion from general TV or video image sequences, there 
are two complicated subjects which have to be considered. One is image processing, for example, automated 
recognition of some human feature points such as the head, elbows or knees etc.. The second is how to estimate the 
photogrammetric camera orientation parameters. 
The authors analyzed the dynamics of the sprinter Carl Lewis /1/ by using TV images of the World Championship held 
in Tokyo 1991 as well as the dynamics of boat rowing /2/3/ of the bow and No.2, gold medalists in the Canadian eight 
team at the Barcelona Olympic Games in 1992, also by using TV images. In both cases, the camera orientation 
parameters could be determined by utilizing fixed information in the images such as the goal line and course lines in the 
first case (Fig.1) and the buoys marking the rowing course lines in the second case (Fig.2). Nevertheless, camera 
orientation for sequential images without such information is an important subject which must be investigated in the 
future. In general, there are two cases which have to be considered. 
1) Fixed sensor and moving object or moving sensor and fixed object : In the case of Carl Lewis, the sensor position 
was fixed and the rotation parameters of the sensor were changing as the camera was following the sprinter. 
2) Moving sensor and moving object : In the case of boat rowing, both the camera position and the rotation parameters 
were moving along with the boat. 
The latter case is dependent on an expensive instrument such as a GPS or a gyro. Then, concerning with case 1, this 
paper describes the application of a video theodolite system for the dynamic analysis of human motion with sequential 
images. 
  
  
       
Orientation image(100 meter race) Figure 2 Orientation image(boat rowing) 
# 
Figure 1 
IAPRS, Vol. 30, Part 5W1, ISPRS Intercommission Workshop “From Pixels to Sequences”, Zurich, March 22-24 1995