OCR using Google's Drive API

[Hrishikesh Terdalkar]

A few days ago there was a post regarding OCR software using Drive API for Windows.

I have been using a python script for this purpose on Linux for the past few years. I've turned that into a pip package.

Package: https://pypi.org/project/google-drive-ocr/

$ pip install google_drive_ocr

It will install an executable `google-ocr`

Some use cases,

$ google-ocr -i image.png

$ google-ocr -b image1.png image2.png image3.png

$ google-ocr --image-dir images/ --extension .jpg --workers 4

Core workflow is,

* Upload an image "as a google document" to Google Drive

* Upon doing this, google adds the OCR layer

* Download google document as plain text

A primary difficulty in using this might be setting up a project on Google Cloud platform and downloading a client_secret.json file.

Instructions for which can be found at https://cloud.google.com/genomics/downloading-credentials-for-api-access

Once done, client_secret.json file needs to be provided

in every call. (To avoid this, one can do

$ google-ocr --client-secret /full/path/to/client-secret.json --write-config ~/.gdo.cfg

Once that is done, next time onwards no  --client-secret option is required.

Regards,

-

हृषीकेश

[Mārcis Gasūns]

So guess it has the same engine as http://ocr.sanskritdictionary.com/ has.

[ken p]

How can I drag image here or post copied image here without uploading file?

Paste, drop, or upload your image

http://ocr.sanskritdictionary.com/