afe/osse_MkII/README.piv


The frame grabber saves the images to 640x482 pixel jpeg
files with the name ``frame.<timestamp>.<level><a_or_b>.jpg''.

<timestamp> is an integer that inicates the number of milliseconds
since the operation system clock of the PC running the framegrabber,
and is used to mark the time between frames.  

<level> is an integer indicating the vertical level of the tank at 
which the frame was grabbed. 

<a_or_b> marks wether the frame in the file is the first or second
frame of the pair grabbed at a particular level.


OBTAINING VELOCITY VECTOR FIELDS WITH PIV

Once you have a collection of pairs of images, you can extract
velocity fields from them.  The current setup uses MatPIV 1.6.1
(see <http://www.math.uio.no/~jks/matpiv/>.  You should look at the
Online Tutorial, particularly "An example session".  Yes, you have to
do it the not so easy way.  This entails establishing a coordinate
system and masking out regions outside of the fluid flow you are
trying to examine.

In order to use MatPIV, you will need to be able to manually examine
at least one of the images created by the frame grabber.  In their raw
state they may be too dark and have too little contrast to be useful,
so you will have to use The Gimp or Image Magick to add contrast to
one of them.  Save it as the file ``masker.jpg''.

To establish a coordinate system, you will have to use one of the MatPIV
utilities to mark tagged points in the image field.  To create these points,
run the Perl script ``addcross.pl''.  This uses Image Magick to add crosses to
the image in ``masker.jpg'', saves the resulting image as ``worldco1.jpg'', and
displays the same.  You may have to manipulate the first couple of lines of
``addcross.pl'' that determine the horizontal and vertical positioning of the
crosses to get the desired result -- the goal is to have at least four crosses
in the field of the particles you wish to include in the velocity field -- the
parts that you later leave unmasked. 

In Matlab, enter the command definewoco(filename,'+'); , where filename is the
image saved with the crosses, and follow the directions in the MatPIV
documentation.  You should enter the coordinates given in the addcross.pl
script.  The Matlab command mask('masker.jpg','worldco.mat'); will then let you mask out regions where no PIV is to be performed.  These areas should include the outside of the tank, the center of the annulus, and the shadow of the laser.
1
2	The frame grabber saves the images to 640x482 pixel jpeg
3	files with the name ``frame.<timestamp>.<level><a_or_b>.jpg''.
4
5	<timestamp> is an integer that inicates the number of milliseconds
6	since the operation system clock of the PC running the framegrabber,
7	and is used to mark the time between frames.
8
9	<level> is an integer indicating the vertical level of the tank at
10	which the frame was grabbed.
11
12	<a_or_b> marks wether the frame in the file is the first or second
13	frame of the pair grabbed at a particular level.
14
15
16	OBTAINING VELOCITY VECTOR FIELDS WITH PIV
17
18	Once you have a collection of pairs of images, you can extract
19	velocity fields from them. The current setup uses MatPIV 1.6.1
20	(see <http://www.math.uio.no/~jks/matpiv/>. You should look at the
21	Online Tutorial, particularly "An example session". Yes, you have to
22	do it the not so easy way. This entails establishing a coordinate
23	system and masking out regions outside of the fluid flow you are
24	trying to examine.
25
26	In order to use MatPIV, you will need to be able to manually examine
27	at least one of the images created by the frame grabber. In their raw
28	state they may be too dark and have too little contrast to be useful,
29	so you will have to use The Gimp or Image Magick to add contrast to
30	one of them. Save it as the file ``masker.jpg''.
31
32	To establish a coordinate system, you will have to use one of the MatPIV
33	utilities to mark tagged points in the image field. To create these points,
34	run the Perl script ``addcross.pl''. This uses Image Magick to add crosses to
35	the image in ``masker.jpg'', saves the resulting image as ``worldco1.jpg'', and
36	displays the same. You may have to manipulate the first couple of lines of
37	``addcross.pl'' that determine the horizontal and vertical positioning of the
38	crosses to get the desired result -- the goal is to have at least four crosses
39	in the field of the particles you wish to include in the velocity field -- the
40	parts that you later leave unmasked.
41
42	In Matlab, enter the command definewoco(filename,'+'); , where filename is the
43	image saved with the crosses, and follow the directions in the MatPIV
44	documentation. You should enter the coordinates given in the addcross.pl
45	script. The Matlab command mask('masker.jpg','worldco.mat'); will then let you mask out regions where no PIV is to be performed. These areas should include the outside of the tank, the center of the annulus, and the shadow of the laser.