PDA

View Full Version : Summary of Replies-Webcams in vehicles



mhely46
05-28-2005, 04:13 PM
Thank you to all those who replied to my query about use of webcams in
vehicles.

While I have still not found everything that I’m looking for, all of the
replies provided valuable information and useful directions, for which I’m
very grateful.

I’m currently looking (with the client) at the possibility of installing
either webcams or surveillance cameras within the vehicle cab at
“piggybacking” on the wireless camera network within their depot to obtain
remote, real-time display and recording.

ORIGINAL POST:
Has anyone had experience using webcams for unobtrusive (NOT secretive)
observation of workers in highly confined surroundings for purposes of
postural and behavioural analyses? Application is a work vehicle
environment, so transient road shock/vibration also present. Any advice,
recommended makes/models, specifications, limitations, things to watch out
for, etc. would be appreciated.

The video would be collected on-the-job, hence the question about webcams
rather than more conventional video apparatus. Am interested in any other
alternatives. Any info on integrating two (or more) simultaneous video
sources to increase accuracy of observations also of interest. Envisage
recording task segments that may range from 1 to 10 minutes in duration.

----------------------------------------------------------------------------
------------------------------

From: Kevin Purdy
K.J.Purdy@lboro.ac.uk

Not quite, but I have used video cameras in similar scenarios combined with
a Sony video walkman (a mini DV recorder/player). These are low voltage and
will work off batteries or from low-voltage accessory regulator(s) from the
vehicle's own power supply. I have used an expensive high quality 'pencil'
camera for 'presentation quality' results, with a separate processor unit
part way down its lead (the particular model I have used is no longer
available), but for scientific/monitoring purposes I've had more than
adequate results from cheep 'board' and 'bullet' cameras from Maplin
electronics (http://www.maplin.co.uk/ )I think these are the analogue video
equivalent to your average video conferencing camera.

The bullet cameras are more robust and easier to place, but the board
cameras can take variety of interchangeable wide angle lenses, which is very
useful in tight spaces. There are even some miniature cmos cameras about 1"
cubed that can have pinhole lenses for the ultimate in discretion and these
are very useful too. Changing light levels as drivers move between sunlight
and shadow can be a problem, as the cameras take time to auto-adjust their
sensitivity and excessively dark or bright conditions are a bit of a
challenge for them sometimes.

For combining pictures I've used a quad video processor - it combines 4
analogue video inputs onto one output (usually in security applications) and
appears as a screen split into four quadrants, one in each corner. These are
also available from Maplins amongst other places, in either B&W or colour
versions. The only downside that I've noticed with the one I've used is that
it seems to lower the frame rate (presumably to a quarter) of any one
picture, which has not normally been a problem unless timing high speed or
short duration actions from the video. I seem to remember this processor was
also a low voltage device, normally run from a mains adaptor but it could
use a regulator from the vehicle's supply.

Incidentally, if you need to run a mains appliance in a vehicle you can get
portable power supplies (some work from batteries and some more noisy ones
work from generators) or alternatively you can just power a device called an
'inverter' from the cars electrical system. Note that some invertors just
give a high voltage DC output suitable for switch mode (e.g. computer) power
supplies, but others give a proper AC mains output that is a lot more
compatible.

----------------------------------------------------------------------------
------------------------------

From: Young-Hui Chang
yh.chang@ap.gatech.edu

I use a Unibrain Fire-i Firewire Webcam with my Vicon system as a visual
check on the data. I don't do any digitizing with it. There is no zoom
feature, but there is a fine focus and the resolution is decent (640x480).
It uses a firewire connection, which
also makes it handy. It sells for about $75, so it may be worth a look.

----------------------------------------------------------------------------
------------------------------

From: Leo
tw129@msstate.edu

I have done similar things before. We did an experiment to video record
driver behavior and response in a SUV. We installed 4 regular surveillance
camera on the car, and a 4-channel synchronized video recording box + video
capture system on the car. The camera is quite small, we have one for hand,
one for feet, one for head and one for the front road view. It worked well.
If that's what you want, Please let me know, I can give you more information
about the model and more specification.

----------------------------------------------------------------------------
------------------------------

From: Mac Reynolds
ERL, LLC
www.erlllc.com
reynolds@erlllc.com

We have used four video cameras in a sedan and SUV to quantitatively measure
posture and position of driver and passenger during 1 to 2 hour highway
drives. The results serve, in part, for the basis of our human body model
and seat design software that is described at www.erlllc.com.

Papers that describe the technology and research are as follows: Stockman,
G., Chen, J. L., Cui, Y. and Reynolds, H. "Measuring Body Points on
Automobile Drivers using Multiple Cameras." Image and Vision Computing
15(4):317-329, 1997.

Brodeur, R. R., Cui, Y., and Reynolds, H. M. "Locating the Pelvis in the
Seated Automobile Driver." SAE 960481. In: Automotive Design Advancements
in Human Factors: Improving Drivers' Comfort and Performance (SP-1155), pp.
41-48. SAE, Warrendale, PA. 1996.

Reynolds, H. M., Rayes, K., Eppler, M., Neal, D., and Kerr, R. "Development
of a Vehicle Laboratory to Investigate Driver Comfort from Physical
Measurements." SAE 960480. In: Automotive Design Advancements in Human
Factors: Improving Drivers' Comfort and Performance (SP-1155), pp. 37-40.
SAE, Warrendale, PA. 1996.

I will be pleased to discuss through email or phone our experience in this
research. We conducted approximately 400 hours of driving studies on the
highway with the system described in the above papers. The results were
extremely informative. It is my conclusion that the only way to know what
people do in their workspace is to record them with video images.

We also used a full-length VCR recording of the drive and found that there
was actually too much data on the recording to study in a large sample. It
is very interesting information and a Master's thesis was written by Maria
Eppler on the topic of analyzing "movement patterns" in the drivers.
[Maria's thesis was submitted at Michigan State Uni., East Lansing,
Michigan]

----------------------------------------------------------------------------
------------------------------

From: Simon Gray
Cambridge Electronic Design Ltd.
WWW.CED.CO.UK --
simong@ced.co.uk

We supply a data acquisition system that records physiological data but it
can also accept video and sound from up to four sources via USB as long as
the web cam(s) are direct show compliant.

The reason for responding to your posting is that the amount of data that
you can accumulate over the 10 minute time frame can be huge. We have found
that using web cams means that there is a significant amount of time after
the recording for compression to finish. This may not be a problem. Some of
the better the cameras and of course the more expensive have hardware
compression which does take the pressure off the PC when streaming. If the
video is to be displayed on the PC while capturing then it has to decompress
on the fly which means the burden is back on. Try reducing the image size as
much as possible!

P. S. We are always interested to hear of published papers mentioning CED if
you have any that you would like to forward to me I would very much
appreciate it.

----------------------------------------------------------------------------
------------------------------

From: Pepe Marlow
Pepe Marlow & Assoc
pepe@netspace.net.au

I cannot give you makes and models of cameras. However Australian motor
racing uses onboard cameras to film drivers in action. You could contact
one of the teams like holden/ford or the television stations to find out
details of camera. I am sure the cameras would be very expensive, but if it
is between rounds you might be able to hire one. Alternatively it would
give you an idea of good spots for positioning a camera.

----------------------------------------------------------------------------
------------------------------
From: Louise Whitby
LouiseWhitby@bigpond.com

There is a company in Qld called Ozspy which I have used for keyloggers
[keystroke data collector for DOS program]. They are very helpful, although
their main business is security and investigations, not OHS.

They have a web site and have wireless cameras can record directly to DVDs
[with the appropriate equipment of course]. Cameras are relatively cheap
and can operate in low light if this is an issue. They are also unobtrusive
but I don't know how good they are up close i.e. in a cabin.

----------------------------------------------------------------------------
------------------------------

From: Tammy Eger
School of Human Kinetics, Laurentian University
teger@laurentian.ca
www.humankinetics.laurentian.ca

We are currently using 3 miniature digital video cameras to record the
posture of an LHD operator. (LHD is a load haul dump vehicle used in
underground mining operations). We are using cameras which are capable of
filming in low-light due to the dark environment present underground.

Mounting: We had metal brackets fabricated to enable us to secure our
cameras directly to the inside corners of the cab. The brackets were
designed to work with existing bolts/screws that were already threaded. ie.
the bolts/screws were already present so we were not altering the integrity
of the vehicle. In order to enable us to move our cameras into the best
position to capture the movements of the operator we attached the cameras to
the fabricated brackets with a ball-joint type mount. The mount enabled us
to move and rotate the cameras into the best position.

Other Details: we also put reflective markers/tape on the operator - we had
tape at the shoulders, on top of the head, and down the back. We were NOT
using the tape to digitize. The tape helped us to view neck, should and
trunk rotation (remember we were in a dark environment).

Equipment Specs:
Cameras:
SONY DCR-PC109 Digital Video Camera
http://www.camcorderinfo.com/content/Sony-DCR-PC109-Camcorder-Review.htm

Mount:
Manfrotto Mini Head 484 (Product code)
http://www.manfrotto.com/product/templates/templates.php3?sectionid=8&itemid
=1949

I have attached a few pictures to try to show the set-up. Our set-up has
enabled us to see the head and truck movements of the operator. If you want
to see more of the operator in the cab you could look into purchasing wide
angle lenses for your cameras.

----------------------------------------------------------------------------
------------------------------
From: Leon Straker
L.Straker@curtin.edu.au

We are planning to use webcams for long duration posture assessment - but
haven't got to technical details yet - we will try to do time lapse so
sampling every 1 second etc. Will be interested in anything you get back.

----------------------------------------------------------------------------
------------------------------

Thank you again, and regards to all,

Max.
__________________________________

Max Hely
Safety Science Associates Pty Ltd
PO Box 2444
Bowral NSW 2576
Phone: 02-4862-1129
Fax: 02-4862-1149
Mobile: 0412-920-300
Email: max.hely@safetyscience.com

__________________________________


--
No virus found in this outgoing message.
Checked by AVG Anti-Virus.
Version: 7.0.322 / Virus Database: 267.2.0 - Release Date: 27/05/2005