![](bg4.png)
Figure 6: The relation between BER and
transmission time
3.2 Estimation of Connectable Units
Now we suppose that all remote stations have to
send one image data to the base station. Let Rn and
Tc be the number of remote stations and the average
transmission time per image data including dialling
time, etc., respectively. There happens many
collisions in connection to the channel (or line),
however with or without any collision, we can
roughly estimate that the total time for all image data
to be sent is approximately Tc×Rn.
Let C be a miscellaneous time such as dialling
time. And let
Nave and Lt be the average number
of image data to be sent and the effective (or active)
time for communication between the base station and
remote stations, respectively. Then, among those
parameters and the aforementioned time
T
0
and Tc,
the following relational expressions hold.
Tc = T
0
×(1+
P
b
×
S
)+C (2)
Tc×Nave×Rn ≦Lt (3)
From (3), the number of connectable remote
stations Rn can be calculated. Now we assume that
the bit error rate
P
b
is 1.0×10
-7
and that the
transmission rate is 9600bps as aforementioned. In
this case, for the simplicity, we set each parameter as
follows.
T
0
=45 [sec/image],
S
=256 [Byte] =2048 [bit],
P
b
=1.0 × 10
-7
, C=10[sec], Lt=24[hours/day]=
86400[seconds/day], Nave=24[images/day] per
remote station.
From (2) and (3), we have
Tc =55 (4)
Rn≦Lt /(Tc×Nave)=65.5 (5)
Then, as a rough estimation in this case, we can
tell that it is possible to effectively use remote
stations up to 65 units. If we consider the worse
condition of bit error rate 1.0×10
-4
, we can calculate
that about 56 units of remote stations are effective.
4 CONCLUSION
We have developed a compact wireless remote
monitoring system that is capable of outdoor image
acquisition in all weather and all day. The remote
station has a built-in CPU and a camera, which is
attached by a solar battery for power feeding. The
built-in camera can acquire images using auto focus,
auto iris, lens zooming, and night vision functions. It
can automatically change into an infrared one,
depending on the lightness. The remote station is
usually based on an event driven method, i.e., it runs
according to the wireless sensor’s signal. Since each
wireless sensor transmits its own radio ID signal, the
remote station can control the viewpoint of camera
according to the signal. Because of those functions
and the compactness of the remote station, multi
remote stations can be flexibly installed as a
multivision system in arbitrary configuration.
The base station also can control the remote
station by giving a compulsory command signal, so
it can obtain monitored images periodically or at any
time from the remote stations.
The proposed monitoring system is adopted by
more than 20 local governments in Japan and is
currently running well.
REFERENCES
K.Yamada, et al., 2000. A Parking Lot Monitoring System
Using Image Processing. IEEJ Trans. EIS, Vol.120-C,
No.6, pp.784-790, (2000-6) (in Japanese).
T.Sogo, H.Ishiguro, Mohan M.Trivedi, 2000. Real-Time
Human Tracking System with Multiple Omni-
Directional Vision Sensors. IEICE Trans. D-II, Vol.
J83- D-II, No.12, pp.2567-2577, (2000-12) (in
Japanese).
H.Mori, A.Utsumi, J.Ohya, M.Yachida, R.Nakatsu, 2001.
Human Motion Tracking Using Non-synchronous
Multiple Observations. IEICE Trans. D-II,Vol.J84-D-
II, No.1, pp102-110, (2001-1) (in Japanese).
H.Nagahara, Y.Yagi, M.Yachida, 2001. Resolution
Improving Method from Multi-Focal Omnidirectional
Images. IEICE Trans. D-II , Vol.J84-D-II , No.8,
pp1882-1890, (2001-8) (in Japanese).
30
35
40
45
50
55
60
65
70
75
80
1.E-08 1.E-07 1.E-06 1.E-05 1.E-04 1.E-03
BER
Transm ittion time[sec]
Send A C K per packet.
Send A C K w hole data received.
WIRELESS REMOTE MONITORING SYSTEM WITH FLEXIBLY CONFIGURABLE MULTIVISION
487