I guess one way is to extract the data in ASCII code and then check the pixel coordinates of where the desired boundary lies in the first image that I have imported. Then I should isolate only these points from the whole image...
Thank you myself.
If someone else has some other idea that could be done within the tool please let me know.
I think my case is quite simpler. It will be the cartesian grid or a pixel valued grid however we want to call it but I need only the velocity of the boundary which is moving and not the rest of the picture which is calculated along with it.
So I think extracting the velocity vectors for every pixel and keeping the information of those that my boundary overlaps with may work.
I haven't tested it yet but this will be my approach.
Thank you for your response, what are your thoughts on the working principle I described?
Στις Δευ, 24 Σεπ 2018 στις 12:19 μ.μ., ο/η William [via PIVlab - Time-Resolved Digital Particle Image Velocimetry Tool for MATLAB] <[hidden email]> έγραψε:
Do you want to measure velocities that are not on a global cartesian grid, but on a grid that is fixed to a moving object?
Just imagine the case of a balloon expanding.
I am only interested in the velocity of the boundary of the balloon for my data.
It seems to me that there are other areas in my image that are calculated as well since the way the light falls on the balloon changes the value of the pixels. So I binarized the image to have a black balloon and a white surrounding area in order to calculate only the velocity of the boundary.
However the software will return the vectors for other pixels as well. So I believe if I take all the velocity vectors for every pixel coordinate and then keep only the calculated velocity for the overlapping pixels with this boundary may be a good approach.
Binarizing is often a problem for PIV analysis, as PIV is based on textures and a BW image does usually not have textures. If your background is not moving, then you can just eliminate all vectors below e.g. 0.01 m/s.