I’m working on refining my app’s processes, and there’s one thing I’d put a pin in a while back that I’m revisiting, and I can’t quite find a good solution, so I thought I’d reach out and see if anyone has ideas.
Basically, I’d like to to be able to SCAN receipts using Adalo, instead of how I currently am doing it, which is just taking photos.
Obviously, Adalo does not have a scanner component, so I’ve been trying to find an API or other service that might allow a workaround. My thought was that users could just take a picture of a receipt using the built-in camera function, then I’d have an API via Zapier or Custom Action take the image file, convert it to a document scan, then re-save the new file in place of the image in the Adalo database.
However, this is turning out to be harder than it sounds. I haven’t been able to find an API that will do that task (I don’t just want to “convert to PDF”, I want it to process the image as a scan, i.e. cropping out background, fixing perspective, enhancing contrast, etc., you know, the way a scanner app would process an image). Furthermore, even if I did find an API that would do the job, I’m not sure I’d be able to get it to work.
Has anyone accomplished putting a scanner tool into their app? Or found a workaround/API that works well with Adalo? Thanks in advance!
Unfortunately I can’t share an exact tutorial (haven’t done this by myself), but if I had such task I’d explore the following path:
use Integromat (Make) to create the whole workflow
you’ll need some file storage to store the results (it’s not possible to upload the file to Adalo collection via API). Though you can store the link to the file
googling by “receipt scanning api” returns quite a lot of results (taggun, lucidtech.ai, ocr.space, …). Though some research will be required to choose the best one.
Thanks for the input. I’ve actually already implemented OCR through an API, so the app is processing receipt data and storing it in Fields in Adalo.
Im more concerned about Image processing, specifically pre-processing the images BEFORE they are sent to the receipt OCR API.
Plain pictures of receipts are much harder for the OCR API to return high confidence results on. If the receipt images could be processed into “scans” first, (again, to crop out background, fix perspective, enhance contrast, etc) it would make the OCR much more reliable AND it would reduce the amount of storage space our receipt images are taking up.
I understand this might be difficult or impossible if Adalo doesn’t just have a scanner utility/component “built in”. But, I also feel like this is a use case that HAS to be common - businesses are increasingly foregoing keeping and filing paper receipts, so it’s natural as this no-code platform develops, more and more users will need to have document scanning functionality for a variety of uses.
Well, as of now Adalo doesn’t have built-in scanner functionality and I’ve never heard of any custom “scanner” component. Could be a good idea for some component developer to create this one (though I guess it should be financially viable).
Regarding APIs - I believe (based on a quick look at their docs) that they do some image preprocessing as well, to optimise specifically for receipts. Though of course more exploration is needed.