OCR detection with XY Coordinates

[Fabian_Weller]

Using the Google Cloud Vision API and Integromat we are able to extract Text / Number from an image and add XY Pins automatically.

You will get something like this without any manual work:

How it’s done:

I would like to use this directly in AppSheet.

Adding @tony @Grant_Stead @Bahbus @tsuji_koichi @Rich

[Grant_Stead]

That’s awesome!

You know, I have a solution that I need to implement this for. I should try and get it done this week. @Martin_Pace feel like helping me get this put together. AI fun!

If I run into any issues I might need your help.

I’ve been taking with @tony about this for the last few months.

[Fabian_Weller]

I’m happy to help @Grant_Stead but I’m sure that I could learn a lot also from you. So please feel free to share what you have found out. You can send it also to wapplabor@gmail.com

[tony1]

@Fabian That’s a really cool demo! I will circulate this with the team.

[Fabian_Weller]

Thank you @tony

By the way: the exact same thing is also possible with PDF files, using the Detect text in files (PDF/TIFF) API which is also available in Integromat.

Unfortunately when using the DOCUMENT_TEXT_DETECTION type, you can only get the full text per-page, or the individual symbols. (not every single word, but every single character)
That’s why you have to put together again all the symbols into words.
It’s doable, but requires a bit more work.

stackoverflow.com

Get Lines and Paragraphs, not symbols from Google Vision API OCR on PDF

python, google-cloud-platform, google-cloud-vision

answered by Dustin Ingram on 09:27PM - 29 Aug 18 UTC

Another solution in Integromat would be:

1. Convert the PDF into JPG (you can convert also multi page PDF files!)
2. Extract the OCR with the Detect text in images API like in the first Post

[Josh_Klassen]

Hi, I would like to watch the video that is in this post but I can't seem to figure out how.

[Fabian_Weller]

Thank you for reporting this @Josh_Klassen 
Here is the video link:
https://www.loom.com/share/2bbdff7349994b6fb0dce72f2234d0ac?sid=82fa2d73-aed3-4231-89dc-d1bc5634b7ee
I've updated also the OP.