The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences. Vol. XXXVII. Part Bl. Beijing 2008
1074
Fig.7 the small object model
Fig.8 One high precision 3D control field
3.2 Experiments
Fig.4 The forth four images of the model plane.
They are photographed with the object distance
further than the former.
At the same time, we photographed a 3D control field with the
same focal length so than we can compare the calibration
precision of the plane calibration and the 3D calibration. And
we make three-dimensional calibration result as a reference
standard.
Fig.5 one image of the model plane,
where the model is not in the centre of the image.
In order to analyze the influence factors, we designed several
experiments.
First, in order to analyze the influence of the camera model, we
take three coefficients into account: the skewness, the ratio of
focal length in the two axises and the distortion coefficients. In
this experiment, we use the images in fig. 1. There is a table
which can descript the result.
In the table 1, we can get that: the coefficients: y, k3 and the
ratio of the two focal length, effect on the calibration result
slightly in the camera, and the appropriate model can get the
right result quickly. In fact, we can choose approximate camera
model without experiment one by one. We can choose the
coefficient by significant analysis, which we won’t introduce
here.
experiment 1
experiment 2
experiment 3
experiment 4
experiment 3
f*= fy, Y=0
fx= fy, Y * 0
f* * fy, Y=0
fx=fy, y=0
f x ^ fy, Y=0
gr
Il
o
k 3 =0
k 3 =0
k 3 ^ 0
k 3 =0
f*
4093.1739
4094.4707
4093.5196
4093.2611
4093.2611
fy
4093.1739
4094.4707
4093. 1395
4093.3723
4093.3723
Uo
1593.6447
1624.5627
1624.2831
1593.4418
1593.4418
Vo
1199.3241
1193.4379
1193.2168
1199.5787
1199.5787
ki
-0.1196
-0.1428
-0.1276
-0.1211
-0.1211
k 2
0.0233
0.3238
0.0667
0.0347
0.0347
k 3
0
0
0
0
0
Pi
-0.0011
-0.1273
-0.0011
-0.0021
-0.0021