Sitemap
Nerd For Tech

NFT is an Educational Media House. Our mission is to bring the invaluable knowledge and experiences of experts from all over the world to the novice. To know more about us, visit https://www.nerdfortech.org/.

Coffee Data Science

Variability in Coffee Particle Distribution Measurement

Qualifying algorithm performance

3 min readMay 14, 2021

--

I wanted to know how well coffee particles could be measured repeatably. The method still doesn’t have a great performance compared to a laser or sifting technique mainly due to estimating volume from a 2D image. However, if it can provide consistent results, then it is a useful tool for me to understand how changes in grind distribution affects the shot.

To examine variability, I took a few samples of the same grind, and I compared them to each other. For each one, I took two photos: normal and zoomed in. I used the normal image with the paper to compute pixels/mm, and then I scaled this factor knowing how much I zoomed in.

All image by author

I usually look at the zoomed-in image to get more detail on the fines (<100um). In this view, the distributions seem to align well except for one sample which seems to have a huge spike at 400um diameter.

So I looked at the normal view to verify because for the normal view, 400um is easily resolved (it’s less than 200um that is tricky for that view). In this view, the distributions line up very well, and the blimp isn’t in the first sample, which suggests it was an error in the zoomed in photo.

It should be noted that there is a high spike at the lowest bin which could be a threshold issue for the image. However, it is consistent across samples, so with respect to understanding variability, it is shows a consistency in measurement.

There is some varibility in the distributions, but the amount of variability is low enough for me to have confidence using a single sample per coffee grind when evaluating grind distributions.

--

--

Nerd For Tech
Nerd For Tech

Published in Nerd For Tech

NFT is an Educational Media House. Our mission is to bring the invaluable knowledge and experiences of experts from all over the world to the novice. To know more about us, visit https://www.nerdfortech.org/.

Robert McKeon Aloe
Robert McKeon Aloe

Written by Robert McKeon Aloe

I’m in love with my Wife, my Kids, Espresso, Data Science, tomatoes, cooking, engineering, talking, family, Paris, and Italy, not necessarily in that order.

No responses yet