Figure 8 shows the left and right image for camera calibration 
which were interpolated by above concept. 
Right image Left image 
Figure 8 Left and Right image 
For the unknown parameters, exterior orientation parameters! <*>o, 
<Po, ko (rotation parameters), Xo, Yo, Zo (camera positions)} and 
interior orientation parameters {/, xo,yo (principal points), a,, a 2 
(scale factor), p 2 (lens distortion)}, camera calibration was 
performed by the bundle adjustment using 9control points. 
Table 2 shows the calibration results for stereo image. 
Table 2 Calibration results 
Left camera 
Xo (mm) 
Yo (mm) 
Zo (mm) 
CO <t> K 
30.248 
109.920 
2382.408 
-0ÌXH0” 0°01’07" o°oo’oo” 
xo (pixel) 
yo (pixel) 
f (mm) 
ai a 2 Pi(H)- 4 * 6 ) 
255.999 
255.989 
17.392 
134.708 0.277 0.10 
Right camera 
Xo (mm) 
Yo (mm) 
Zo (mm) 
co 
<t> 
K 
246.126 
113.171 
2382.831 
0°01 12” 
0°01 58” 
o°oo’oo” 
xo (pixel) 
yo (pixel) 
f (mm) 
ai 
a 2 
Pi (10*) 
256.004 
255.989 
16.927 
134.560 
0.120 
-0.10 
4 AUTO-TRACKING 
After the camera calibration, the video theodolite tracked the test 
model was slowly moving to left side. Image A2 was taken at 
+250mm. The changing values of the target on the actuator were 
then controlled through a personal computer. The rotation angles 
of the video theodolite were 5°52 48 clockwise, 0°04 13 under 
the horizon, thus giving zlFj=0°04 13 and AHj=-5°5248 . 
Image A3 was taken at +500mm and ZjFj=-0°12 48 
-12°03 51 . Similarly, image A4 and A5 were taken at -250mm, 
-500mm on the right side, respectively Z]F<—-0°25 15”, A\H(= 
8°18 5l” and ZH / i =-0°26’l4”, AHs=\4°\3 29. 
The unknown parameters, co and <fi for stereo image of each 
sequential image should be estimated as the sum of changing 
vertical and horizontal values resulting in <y o and 4> o 
respectively. Here, coo and 4> o are the calibration results of the 
both orientation image. Consequently, co and <p are calculated 
as follows using the changing values in vertical (AV), horizontal 
direction(ZlHi) and coo, <t>o- 
However, each sequential image was taken at a different 
exposure station, due to the discord between the center of the 
theodolite and the lens of the CCD 2 camera. Each camera 
position has to be corrected to respond to the rotation of the video 
theodolite by the following equation, 
Xo' =D*cosVsin(<p -Ho) 
Yo =D {cos co sin A - sin cocos A cos( <{> - Ho)} ^ 
Z =D* {sin co sin A+cos cocos A cos( <t> - Ho)} 
Zo - (D-Z')cos Vo 
Where, Xo', Yo', Zo', corrected camera position, 
A=tan I ( Yo/ (D-Zq) ) d' = yjxl +T 0 2 +(D - Z Q ) 2 
Auto-tracking test procedures are shown in Figure 9. 
Tracking 
Calibration 
/ Color information 
Extraction of color 
Binarization 
Labeling 
Area gravity 
Image Processing procedure 
7 //, V angle 
Calculation of Ground Coordinates 
Calculation of Velocity 
[ Superimpose L 
(•Ground coordinates (X, Y, Z) H 
I •Velocity 1 
[ Calculation of Rotation Angles \ 
Control of Rotation Speed 
I No L \ W 
End 
Figure 9 Tracking procedures 
The detail procedures for the auto-tracking test are as follows: 
+Window is previously cut out and color information for the 
marker is extracted so that the video theodolite can be tracked the 
object. 
+ Area gravity of the marker for the both odd and even field is 
calculated using binarization and labeling procedure. 
+Three-dimensional ground coordinates for the area gravity can 
be calculated using the calibration parameters and rotation angles. 
Since the unknown rotation parameters, co and <t> for each 
stereo image can be obtained as the sum of changing vertical and 
horizontal values resulting in co o and <p o for the orientation image 
respectively and parameters other than co and <f> are considered 
as the same values as the calibration results for the orientation 
image. 
co - cooX AtVi </>-<?îoXAfri 
(1)