1. Receive images from field and dump JPG (or process raw to JPG) into folder
2. Users given a link to click that would give them a page with the clickable image and set of classes (species) to ID
3. 'Next' button to present next image, and save JSON or CSV of the data clicked at same time.
I'm trying to solve the same problem as Ragnvald (OP) - while CVAT and some other apps get close to what we need, I have a feeling that building on a super-simple application as you demonstrated would be productive
I'm looking at your Github repo for this, digging into the code a little
Saving the coordinate JSON or CSV on the local PC would probably be easier, on the net somewhere would need some sort of storage backend, probably harder.
Ragnvald - I'm thinking this would be more for an annotation lab, a replacement for the very heavy CVAT or VOTT apps, but not usable on smartphone. I suspect a specific app would be needed for that.
@til_b if the image is loaded directly, not using an OSM map as your example has, will it still be zoomable?
I can ask an ex-colleague of mine, but I don't know how proficient in leaflet and JS he is. I think he could manage, though. I haven't talked to him lately so don
Good if you can check this out. Perhaps I and Simbamangu in any case should start writing a requirement/specifications document. How would that sound @Simbamangu ?
Done, invite for Google Doc sent to you @ragnvald. @til_b would need a contact for sharing GDoc, not sure there's a way to privately message here on SE.com?
Example looks like something which could work - with some polishing on the user interface. Automation of image sources as well as option to send/store data.