Experimental Physics and
Industrial Control System

Emmanuel Mayssat <[email protected]> · Wed, 29 Aug 2007 12:28:11 -0700

All,

Following up on an old thread, I moved forward with an axis video
server.
I noted that Motion JPEG is the image format that is being used.
To grab a snapshot we can use a program called axmjpeg that can extract
snapshot from a video stream. This script works quite well. Initially I
was afraid that disk write (of snapshot files) would significantly slow
down the snapshot acquisition. But it doesn't appear to be the case (for
10Hz acquisition).
(Note that the axmjpeg program can pipe its output as well (no disk
write), or can directly be integrated in an epics driver (that's the
beauty of having the source code)).

Then I closely looked at the JPEG format.
JPEG is well suited for video IP (for compression reasons), but that the
losses (due to compression also) may impact computer image analysis.

When describing the JPEG compression algorithm:
"The reason for doing this is that you can afford to lose a lot more
information in the chrominance components than you can in the luminance
component: the human eye is not as sensitive to high-frequency chroma
info as it is to high-frequency luminance."
So JPEG is no good for images with high transition in the chrominance
alpha channel. 

Ok, keeping this in mind, I decided to move forward with JPEG.
No I could use my custom library to compute beam profile, position, etc.
But is there a standard package for image analysis I should know about?
Or is everybody doing it there own way?

Regards,
--
Emmanuel

On Tue, 2007-04-10 at 13:48 +0100, Pearson, MR (Matthew) wrote:

> doesn't have a web server, and needs windows software to view the video.
> There are software projects out there which can capture IP video, but
> not sure what Linux support exists.

Experimental Physics and Industrial Control System

Experimental Physics and
Industrial Control System