http://imagej.273.s1.nabble.com/detect-motion-across-frames-tp3696278p3696292.html
Gary.
> I have images that resemble clouds and I am also trying to track the
> motion of clouds from frame-to-frame. Although my "clouds" are actually
> clusters of small particles.
>
> I've been considering using cross-correlation techniques. I know they
> will work because I have large structures that move small displacements
> from frame-to-frame. Cross-correlating will give the magnitude and
> direction of the displacement of the structure between frames.
>
> Does anyone know if ImageJ has the capability to use cross-correlations to
> measure the displacement of large structures from frame-to-frame?
>
> I'm a new user to ImageJ and I continue to be very impressed. I was going
> to buy a commercial image processing software package for $5000, but I
> find ImageJ to be better! I will be adding plugins to ImageJ in the near
> future.
>
> Frank Shaffer
> USDOE
>
>>>> "Brian Willkie" <
[hidden email]> 4/7/2008 2:38 PM >>>
> Hello,
>
> I searched the list archives briefly, but didn't find this topic
> specifically. It's entirely possible that I simply didn't know what to
> look for.
>
> I have access to GOES 12 satellite data (i.e. cloud data) as either
> raw data, TIFF files - frame by frame - or High definition QuickTime
> movie.
>
> I'd like to use some image-processing techniques to identify pixels
> that you and I would group togther as a clouds, and track their motion
> from one frame to the next.
>
> Analyze particles seems to work well to identify clouds in a given
> frame. Are there any commands / plug-ins / algorithms to track their
> motion/change across frames (i.e. can I get a dataset that tells me
> that a given particle in the current frame is the same particle as one
> found in the previous frame, although its size, shape, and/or position
> have changed slightly)?
>
> Thanks,
> Brian
>