After many rewrites I'm still not entirely thrilled with how I've presented this.... please delete if inappropriate.
As part of my job, we record observations on underwater paper and then have to enter the data manually at the end of the day, a time consuming and error prone process
We have set forms that we write data into as seen in the attached image
Manually entering the data for the top two rows isn't a problem (reef, site, date, observer, slope and benthic cover)
However, manually entering the data for each of the categories e.g Acropora, isopora etc is problematic.
The sheet shows the genus of coral on the left and what transect it was counted on.
What I would like to do
I have been thinking about automating the entry process and so far have been looking at using pytorch to train an algorithm to count the tally marks and write it to a csv
I have quite a large training set that I could use
I don't have much of a background in python but am quite comfortable in R as I use it most days for work and was thinking I could use the reticulate package.
There are different forms for other tasks that we also manually enter but I plan to carry the method over to the other sheets
I haven't found an off the shelf software package to do it so I assume I'll have to write something
So far, I'm thinking I'm going to approach this by first identifying each cell ie T1, Acropora then the machine counts the tally marks in that cell.
Assuming that is a practical way to approach it, what would you call that process so that I can research it and find a way to script it?
I'll take any suggestions on how to approach this :)