OCR From Screen Data

I would like to extract data from an internal application. The data is presented on screen as an image, so I want to use OCR to get the data.

I can achieve the result if I have a static image (i.e. save the image and OCR that). However, the automation I am generating will not always have the same data on screen (the account names are all different) - but it will always be in the same place.

So my requirements are:

  1. set up a loop to check many accounts (I have a CSV file with all the account numbers - the loop works correctly)
  2. automate the keystrokes to open a screen based on an account number (working)
  3. the system displays the results for that account (Working)
  4. Use OCR to read the account name displayed on screen (This is where I have an issue)

There are loads of extra steps needed, but I am looking for information on how to perform step 4 given that the actual image is different each time.

Thanks in advance.

Hi Colin,
The part of the image that you use as the anchor area has to be the same in all images. Please share a screenshot of your ocr action to see the anchor and ocr areas.

Thank you for getting back to me so quickly.
It turns out that if a version of the application is already open, the second version is opening below the other, which means the anchor is pointing to the wrong location. I’ve added a check to make sure the application is always closed before the automation starts.
Thanks for your help,

1 Like