New version - major beta release is on its way

We added so much functionality during the last weekend sprint on URAPIV_GUI:

1. due to a request of our users at the University of Ben Gurion, Beer Sheva, Israel, we added an option to read MetaMorph TIFF images, using the tiffread function of Francois Nedelec, EMBL, Copyright 1999-2006. [http://www.cytosim.org/other/tiffread.m]
2. we added an option to use RECTANGULAR interrogation windows (64 x 16 pixels, etc.)
3. we added Select Region-of-Interest and Reset Region-of-Interest - smaller windows are available for quick-n-dirty check of your images
4. we added an option (to GUI, lost functionality due to transfer to ETH format of high-speed PIV) to read images named as:
   • bird001a.tif - bird001b.tif
   • bird001_b.tif - bird001_c.tif
   • bird001.tif, bird002.tif, bird00x.tif
5. New look:

 

(press to see the full image)

.:: January 05, 2008 ::.

New Wiki on the Sourceforge.net

Don’t worry, we will not drop the WordPress blog (which is great) and we like it a lot, but since Sourceforge.net are also progressing, we have added two major features:

1. Subversion control (finally :-))

http://matlabpiv.svn.sourceforge.net/viewvc/matlabpiv/

2. Wiki

http://matlabpiv.wiki.sourceforge.net/Getting+started 

Enjoy it

Note the updated links on the right >>>

Note that we’ve updated the links to the personal websites of the group members.

URAPIV in dusty plasmas (including the tutorial)

The story started from one e-mail:

I would like PIV to evaluate some of my video data. In search of a solution, I have found your tool, URAPIV. I have downloaded URAPIV_GUI_TIFF (Release Date September 20, 2004) from sourceforge.net.Unfortunately, I am running into a few problems. First, I am only able to read in 4 images at a time. In addition, the code only seems to work if the images that are included with the URAPIV_GUI_TIFF download are included in the directory containing the images that I would like to analyze.

Do you have any ideas on how to have the software read in all of the images? Also, is there a manual for this software?

Many thanks and best greetings.

Jeremiah Williams

So, we answered (as usual), suggesting to take a first look at the images and to reproduce the problem (that sometimes happen simply due to Matlab version, compatability issues, etc.) and got a sample of 10 frame AVI file. Since it was not our first time to get the AVI files, we decided to take a step and slightly adopted the URAPIV code for reading and working with AVI files. The technical details are in the Tutorial. Here we’d like to tell you the story behind the images. Our users come from the well known (in plasma physics and beyond it) lab from Auburn University:

Particularly, we have found the old-dated pages about PIV analysis in dusty (or complex) plasma, e.g.

The lab knows about 2D PIV for a long time, as one can see from the short intro to PIV on their web site:

And it’s an honour for the URAPIV project to be helpful in this research. The output of our short customized solution is an updated URAPIV_AVI version (that’s recently has been uploaded to the sourceforge.net) and the example looks like:

Read more about the technical stuff on the Tutorial page.

The proper credit should be as the author says:

The video comes from the Colloidal (Dusty) Plasmas research group at the Institute of Physics, University Greifswald, Germany. The video was shot by Matthias Wolter, who has been visiting our lab for the past couple of months, during the 8th DLR parabolic flight campaign. In the video, a plasma crystal under microgravity conditions was manipulated by a 600 mW laser.
You can find additional information on the type of work that he does at his groups web page (http://www5.physik.uni-greifswald.de/).
The PIV aspect of this work comes from work done at our group, the Plasma Sciences Laboratory at Auburn University. We have applied PIV techniques to the study of complex (dusty) plasmas for a number of years and have two commercial PIV systems (a 2D system from TSI and a stereo system from LaVision). Additional information on this can be found at our group’s web page, http://narn.physics.auburn.edu.

With kind regards,
Jeremiah Williams

URAPIV for cars?

The researchers from the International Automotive Research Centre at the Warwick University has asked if URAPIV can perform on the images like the example above. The original images are typical gray scale images, but taken with a camera that also moves. Therefore, the motion is a superposition of the motion of cars (note the direction of the middle jeep) and of the "optical flow" because of the camera shift. Anyhow, it was an interesting challenge and we were happy with the preliminary result. For this purpose we also updated URAPIV with a version for Matlab R14SP2 (R14 is ver 7) that reads TIFF, JPG, BMP or all other formats that IMREAD function of Matlab can read. The names are like from a high-speed camera (sometimes called time resolved PIV or high frame rate PIV, etc.) and we analyze 1st with the 2nd, 2nd with the 3rd, and so on.

The strange image is an overlay of the two images in one RGB image, when the first one is the red layer (R) and the second image is the blue layer (B). Arrows are green, respectively. The magenta is all the regions that didn't change and the differences are seen between the images where the color is blue or red. In Matlab we do it like this:

>> imshow(cat(3,a,a*0,b)); % a/b are the first/second image  

Don't hesitate to ask for the most recent version by e-mail, we are always glad to see URAPIV that helps to resolve different problems.

Two things about moving between URAPIV and PyPIV (i.e. Matlab and Python)

There are two main sources of information for our users that want to join us in the move from (we stress out that the move is not complete move, we keep URAPIV development and support) URAPIV to PyPIV, or in more general sense, from Matlab to Python.

Very useful set of commands in PDF format by Johnvey Hwang, aka 37 mm website - see the blog here or press on the [link] to download the PDF.
A new Wiki, called NumPy for Matlab Addicts (that we fell like ones). NumPy is a package we use for PyPIV

I want to break free :-)

I want to break free …. (Queen)

URAPIV wants to break free … (of Mathworks licensing policy)

URAPIV now is also* PyPIV, where Py is for Python (http://www.python.org)

* means that we are not going to drop the Matlab toolbox, but in parallel we are going to enter a new field: ABSOLUTELY FREE OPEN SOURCE PIV ANALYSIS (as they say think of "free as beer"). Even kids in a primary school in Gana can do PIV now - it's free as Python programming language is free, it's free as Python Imaging Library is free, it's free as Scientific Python is free.

If you install totally free OS for human beings - Ubuntu (or any other Linux distribution), you have Python built-in

If you have Windows, download an excellent distribution from Enthought Inc. and run the Windows installer. You'll have Python working in a minute.

If you're in Mac OS X, we still have not tested it, but MacPython is there and Fink is there, so it should be straightforward installation.
Happy PIVing to all

Here's a screenshot of PyPIV on a PC with Windows XP and Enthough Python at the background:

Upgrade for Matlab 7 (R14SP3)

We call all our users, who upgraded his Matlab version to R14SP1+ (i.e. SP1, SP2 or SP3, see Mathworks web site for more details) to upgrade his URAPIV.M from the download site: http://sourceforge.net/projects/matlabpiv
URAPIV.M is now a self-inclusive function, that contains all the necessary utility functions (ReadImdir() and QuiverM()) inside, sharing the variables without calls-by-reference. Moreover, ReadImDir() is updated to read more image formats: JPEG, TIFF, BMP (and can easily be extended for more formats) and also reads sequential lists (from High-frame-rate PIV or high-speed PIV or cinematographic PIV) and also from cross-correlation cameras (a usual _b.xxx and _c.xxx convention for the base and cross frames).