The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences. Vol. XXXVII. Part B5. Beijing 2008 
942 
The 2D residual patterns of the matte black target are presented 
in Figure 6. The distances between the targets and the range 
camera for Figures 6a and 6b are approximately 25cm. In 
Figure 6a, two abnormal residual patterns are observed. First, a 
high residual is shown in the boundary of the target, which was 
observed in Figures 4 and 5 in the previous section. Second, in 
the centre of the matte black target, the same amount of the 
residual as the target’s boundary is observed and it is caused by 
the higher returned intensity than the case of the Figures 4 and 5 
since the distance of the camera and the target is shorter. In 
Figure 6b, the residual pattern of the matte white target at the 
distance of 50cm presents the same pattern as the matte black 
target in Figure 6. 
EXPERIMENT II 
As the second part of the experiments for this paper, we 
conducted several precision tests with various hardware setup 
such as warm-up time, integration time and amplitude threshold. 
As an example, the intensity and range images from 
SwissRanger SR-3000 are presented in Figure 7 with the 
distance threshold (minimum distance = 1.80m and maximum 
distance = 2.20m). 
(a) Intensity image (b) Range image 
Figure 7. Intensity and range data from SwissRanger SR-3000. 
1.1 Warm-up time 
For the test on the warm-up time, the distance of the target from 
the camera was approximately lm and the same setup as Figure 
3a with a flat target with grey colour was utilized. A continuous 
measurement was conducted for 8 minutes. The interval 
between each measurement was 1.5 seconds and 20 seconds 
pause after every minute. The estimated distance by the first 
order plane fitting and the mean intensity of the target are 
presented in Figure 8 with respect to the measurement time. 
3400 
I* 3300 
In 
C 
0) 
3200 
-E 3100 
1.100 ^ 
1.000 v 
0.900 c 
0.800 tS 
0:00:00 0:02:53 0:05:46 0:08:38 
Time [hh:mm:ss] 
Intensity 
Distance 
After turning on the range camera, the estimated distance 
increases within the time window of about 2 minutes as shown 
in Figure 8. The approximate accuracy error is about 10cm 
within 2 minutes after the starting of the measurement. Further 
more, a relationship between the measured intensity and the 
accuracy of the range camera as shown in Figure 8. In other 
words, its bias in the distance measurement is directly related to 
the warming up of the camera. From our experiment, the 
minimum warming-up time of 6 minutes is recommended. 
1.2 Repeatability 
Before the metric performance tests with both integration time 
and amplitude threshold of SwissRanger SR-3000, we 
conducted a test on the repeatability or reliability test of the 
range camera. To do this, we conducted a successive 30 
measurements in the different distances of the target, from lm 
to 4m, with interval of lm. The mean estimated distance of the 
30 measurements was utilised to calculated a reference distance, 
named D mean . In Figure 9, the difference between the distance of 
the estimated distance of each measurement and D mean is 
presented. For the cases of the distances of the target from the 
range camera (100cm, 200cm and 300cm), the standard 
deviation of the difference from D mean is within 2mm. In the 
case of 400cm, the standard deviation of the difference from 
Dmean IS about 1 Ctt!. 
Figure 9. Difference from the mean distance (D mean ) with 
various target-camera distances from 100cm to 400cm. 
4.3 Integration time 
In this section, some test results on the integration time will be 
presented, which is one of the most important parameters of a 
range camera such as SwissRanger SR3000. In order to obtain 
knowledge about setting the right or optimal integration time, a 
grey metal target were used. In addition, in order to a reliable 
measurement, a total amount of 30 measurements were 
recorded for every setting and were averaged for the least 
squares adjustment. 
Two plots of 3D point clouds from the grey flat target with two 
different integration times, 20ms and 51ms are presented in 
Figure 10, as an example of the important of the integration 
time for a range camera. Around the centre of the target, we can 
observe some erroneous measurement by the range camera. 
Markers were placed on the ground in a distance of 0.10, 0.25, 
0.50, 0.70, 1.00, 1.25, 1.50, 3.00, 3.50, 4.00, 5.00m measured 
from the back side of SwissRanger SR3000. The target moved 
from one marker to another after every measurement. Note that 
Figure 8. Warm-up time vs. Mean intensity and the estimated 
distance of a flat target from the range camera.