Basic Setup Question

[D Hill]

I am looking to set up a basic webcam (venus USB 2.0 webcam) to take a picture every few minutes over a period of hours, and then calculate the greyscale average of a particular ROI, and I would like this to be as automated as possible. I can find various bits and pieces - for example I am ok doing the ROI, but I am not sure how to tie all this together. Initially this would be done on a Mac, but eventually I will be running it on a PC too. Does anyone know of any particular plugin/script combinations which would be good for this?

thanks.

[Gabriel Landini]

On Wednesday 17 Nov 2010  13:21:58 you wrote:
> I am looking to set up a basic webcam (venus USB 2.0 webcam) to take a
> picture every few minutes over a period of hours, and then calculate the
> greyscale average of a particular ROI, and I would like this to be as
> automated as possible. I can find various bits and pieces - for example I
> am ok doing the ROI, but I am not sure how to tie all this together.
> Initially this would be done on a Mac, but eventually I will be running it
> on a PC too. Does anyone know of any particular plugin/script combinations
> which would be good for this?

Hi,
SambaJ had some interface to capture from a webcam, not sure if that is still
maintained or if it will drive your device.

However, I doubt that using a webcan will produce any usable data. Webcams
output lossy compressed signals and most commonly have automatic gain (in many
this is impossible to turn off). That means that the intensity measurements
will not be controlled (the image gets darker and the camera will "compensate"
for that, changing the gain). Measurements of grey intensity obtained that way
will be meaningless.

Depending on what you want to do, it might be an option to spend some time
interfacing a photocell via an AD converter (there are several cheap options
that use the parallel port or the USB), calibrate it, so you know what you get
in the signal, and use that instead.

Cheers

Gabriel

[Stephan Saalfeld]

Hi,


On Wed, 2010-11-17 at 13:49 +0000, Gabriel Landini wrote:

> On Wednesday 17 Nov 2010  13:21:58 you wrote:
> > I am looking to set up a basic webcam (venus USB 2.0 webcam) to take a
> > picture every few minutes over a period of hours, and then calculate the
> > greyscale average of a particular ROI, and I would like this to be as
> > automated as possible. I can find various bits and pieces - for example I
> > am ok doing the ROI, but I am not sure how to tie all this together.
> > Initially this would be done on a Mac, but eventually I will be running it
> > on a PC too. Does anyone know of any particular plugin/script combinations
> > which would be good for this?
>
> Hi,
> SambaJ had some interface to capture from a webcam, not sure if that is still
> maintained or if it will drive your device.
>
> However, I doubt that using a webcan will produce any usable data. Webcams
> output lossy compressed signals and most commonly have automatic gain (in many
> this is impossible to turn off). That means that the intensity measurements
> will not be controlled (the image gets darker and the camera will "compensate"
> for that, changing the gain). Measurements of grey intensity obtained that way
> will be meaningless.
>

Not if you have a calibration card in each image:

http://www.hawksphotovideo.co.uk/calibration_and_grey_cards/Lastolite_XpoBalance_Notemaker_Calibration_38cm_15_1559

That would allow to compensate for whatever bogus the camera is doing.

Best,
Stephan

[Gabriel Landini]

On Wednesday 17 Nov 2010  15:18:19 you wrote:
> Not if you have a calibration card in each image:
> http://www.hawksphotovideo.co.uk/calibration_and_grey_cards/Lastolite_XpoBa
> lance_Notemaker_Calibration_38cm_15_1559
>
> That would allow to compensate for whatever bogus the camera is doing.

Hi Stephan,
If one can fit that in the frame yes that is a possibility to try and minimise
the autogain settings, but it is not straightforward when using a microscope
(and yes, webcams are being sold with ocular adaptors for microscopes). The
adaptors are great, the cameras are not. :-)

One could create a microscopic version with stepped (overlapping) neutral
density filters... somehow spending thousands in good optics for a microscope,
to capture the data with a £20 webcam somehow does not sound right.

Webcam images have also plenty of compression artifacts, presumably to keep up
with the transfer rate. I am not sure how one would avoid that.

Cheers

Gabriel

[Glen MacDonald-2]

Dear D,
We used an old analog CCD camera (Sony XC-77) with this device to record fish motion for rheotaxis assay.  I think the capture was controlled with BTVPro.  The output was saved as Quicktime without compression and analyzed with ImageJ.  

http://daystar-store.com/xlr8proviewusbfullscreenvideocapturepackageforusb20.aspx
ProView USB - USB Video Capture for Mac & PC

Regards,
Glen

Glen MacDonald
Core for Communication Research
Virginia Merrill Bloedel Hearing Research Center
Box 357923
University of Washington
Seattle, WA 98195-7923  USA
[phone]
[hidden email]








On Nov 17, 2010, at 4:33 AM, D Hill wrote:

> I am looking to set up a basic webcam (venus USB 2.0 webcam) to take a picture every few minutes over a period of hours, and then calculate the greyscale average of a particular ROI, and I would like this to be as automated as possible. I can find various bits and pieces - for example I am ok doing the ROI, but I am not sure how to tie all this together. Initially this would be done on a Mac, but eventually I will be running it on a PC too. Does anyone know of any particular plugin/script combinations which would be good for this?
>
> thanks.