2023-05-02

This is Edouard's DBGI daily open-notebook.

Today is 2023.05.02

To do

Presentation

Show github: https://github.com/digital-botanical-gardens-initiative/iNaturalist_uploader

What the code is doing:

Downloads gpkg layers (point layers) and pictures
Changes pictures names by deleting all problematic symbols and replacing white spaces by underscores
Detects the CRS (Coordinates reference system) of gpkg files and converts to EPSG:4326, that is the CRS used by inat
Converts the gpkg files to csv files
Compresses pictures if they are bigger than 5MB (while loop with little decrease each time, so that the pictures are not compressed to heavily)
Adds coordinates and dbgi code in pictures metadata to facilitate inat import.

Run the code

Share the folder: python3 -m http.server

Access the folder: http://83.77.116.250:8000/

Download a picture and show an example on iNaturalist

Now stays some questions to finish the job correclty:

Which cloud to use? (already discussed about nextcloud)
Where to host ? (On the field collection server or on another server)
clicbot or not clicbot ? And if clicbot, how to manage it? (As inat won't be happy if they know)
How to organize pictures on the cloud? (one folder per sample with folder named with dbgi code, one folder per layer, one folder per layer per upload, ...)

Set up

Create shared folder: python3 -m http.server

To access the sharde folder: http://83.77.116.250:8000/

notes

The code to change pictures metadata is done and is working. It is available on github: https://github.com/digital-botanical-gardens-initiative/iNaturalist_uploader

The code is made to adapt. It signifies that if we create a new project or a new layer, it will be integrated. If a project is created, don't forget to add the dbgi user as manager, because it is the one that is used to download the files.

What the code is doing:

Downloads gpkg layers (point layers) and pictures
Changes pictures names by deleting all problematic symbols and replacing white spaces by underscores
Detects the CRS (Coordinates reference system) of gpkg files and converts to EPSG:4326, that is the CRS used by inat
Converts the gpkg files to csv files
Compresses pictures if they are bigger than 5MB (while loop with little decrease each time, so that the pictures are not compressed to heavily)
Adds coordinates and dbgi code in pictures metadata to facilitate inat import.

Now stays some questions to finish the job correclty:

Which cloud to use? (already discussed about nextcloud)
Where to host ? (On the field collection server or on another server)
clicbot or not clicbot ? And if clicbot, how to manage it? (As inat won't be happy if they know)
How to organize pictures on the cloud? (one folder per sample with folder named with dbgi code, one folder per layer, one folder per layer per upload, ...)

Solutions

Try to make 2 different tags in Subject Homogeneize the pictures sizes (all to have the same pixel sizes) cookie cutter -> makes packages Nextcloud good solution cthoyt.com -> documentation, how to do packages

code

links

test