OCR - Optical Character Recognition (for Custom plans)

If you purchase a custom paid plan with ParseHub you can add on OCR (Optical Character Recognition) which will allow you to scrape text out of images with a minimum accuracy of 90%. You can contact us to learn more about this feature and our custom plans.


I'm using a company profile from HKTDC.com as an example for this tutorial, if you want to follow along.


Opening the "More" button 

The telephone number and fax number can be found on the company profile after clicking the "More" button on the bottom left hand corner.


To open the more button on your project: 

1) Add a select command. 



2) Select the "More" button.



3) Add a click command.



4) Select the option to "Continue executing the current template".



Using OCR to extract text from an image


1) Add a select or relative select command.



2) Select the image you would like to extract text from. 



3) Press the + button next to your select or relative select command and add an extract command.



4) In the text box underneath the Extract dropdown, select the "Read text from image" drop-down option or input $e.ocr



Note that the OCR function will not work on test runs, you will see the following message:

"OCR is disabled in test run and the client. You can inspect your results by running your project through the \"Run\" button in the \"Get Data\" screen." 

Have more questions? Submit request!


Article is closed for comments.