Example images of active matter

[Zhengyang Liu]

Here are some images from the field of active matter that can be analyzed by PIV. We could include them in the examples / tutorials. The idea is to exhibit a broader range of problems that can be solved by the OpenPIV package. 

1. Bacteria in a droplet raw images

2. Active nematics raw images

[Alex Liberzon]

active nematics images have zero displacement - this is probably a mistake

[Zhengyang Liu]

My apologies. I uploaded the images apparently in a rush and did not check carefully. Thanks for pointing out. I will share new images soon! 

[Alex Liberzon]

Here's an example with a Jupyter notebook, using OpenPIV 0.26.0 with static masking

https://github.com/OpenPIV/openpiv-python-examples/blob/main/notebooks/bacteria_in_a_circle.ipynb

[Zhengyang Liu]

Nice example, thank you Alex! 

It seems that the multi-grid window deformation algorithm `windef` is a more preferred method for my images, which I did not use before. I compare below my old method (as described in the first tutorial based on extended_area_search_piv), simple_piv and windef. 

I tried to find more information about windef in the documentation site (readthedocs), but only found a short tutorial (https://openpiv.readthedocs.io/en/latest/src/windef.html), which does not say much about how it works and what advantages to expect. windef also generates many intermediate results (images) which are supposed to be informative of what is going on in the backend, but I don't find a guide about how to use this information. `windef` might be a good place to improve the documentation.

I also notice that when I do the quiver plot using `matplotlib.pyplot.quiver(x, y, u, v)` (left panel), the vertical velocity components all look inverted. As a result, I have to do `quiver(x, y, u, -v)` to make the vector field consistent with the motion (shown below). I might misunderstand the convention, so it would be helpful if someone can help me understand this.

I also notice something I don't understand in the last two output images of windef: the y-axis direction is flipped, from [350, 0] to [0, 350], but  the velocity field remains the same. 

[Zhengyang Liu]

I have also uploaded new images for active nematics at the same link. Hopefully no mistake this time!

[Alex Liberzon]

the code cites a couple of papers that the code is based on - it'll be useful to read it and see if you find there explanations. unfortunately no time to write stuff. it would be great if you could help. 

please read the thesis of  Theo Kaufer   it's linked to the documentation or you can google it. maybe ask here and Theo will send you a copy from which you could help everyone to understand it better. 

Best regards

Alex

[Alex Liberzon]

did you try windef, following the same ideas?

[Zhengyang Liu]

@Alex No not yet. I will follow the example you made for the bacteria image to produce an active nematics counterpart when I have some time. For windef, I'm interested to do more documentation about it since it looks like a useful feature. I tried to google @Theo Kaufer's thesis but was not able to find it. If you can kindly provide me a link to the thesis, as well as the papers you mentioned, I am happy to read through and to write something we can then review together.

[theo.ka...@gmail.com]

Hi,

Here is a link to a paper that describes the method:

https://iopscience.iop.org/article/10.1088/0957-0233/13/1/201

 I have not published the thesis, but I have sent you a private message with the attached thesis.

[Alex Liberzon]

Hi, 

If you're interested, please add the correct way to credit the authors of this data and maybe a link to the lab, group and/or citation to some paper that is linked to this data? 

There are already two demos - the Jupyter notebook that I wrote and the new shiny VIKTOR app https://cloud.viktor.ai/public/particle-image-velocimetry-in-bacteria-demonstration 

plus there're some twitter and linkedin posts that you could use to get more attention to your work. 

regards

Alex

On Monday, March 13, 2023 at 7:26:41 PM UTC+2 liuzy1...@gmail.com wrote:

[Zhengyang Liu]

Hi Alex,

Thanks for making the demos, they are awesome!

The bacterial droplet data were produced by me in the lab of Eric Clement at PMMH, ESPCI Paris. The active nematics data were produced by me in the lab of Teresa Lopez-Leon at Gulliver, ESPCI Paris. There are no papers linked to these data yet, since we are still working on it (PIV is an important aspect of the analysis, and that's why I'm here trying to get involved). I've put links to the lab websites, which can be used to credit the authors of the data.

Best,

Zhengyang