I have a few images which contain Medical information. For example Patient Name, Doctor Name, Date of Birth, Total billing amount. All those information are at an almost fixed part of the image and very little based on how it scanned (image is either in middle or in right or tilted a bit).

Can we have Deep Learning/ Neural Network solution where it can identify that certain location has only Patient Name or Doctor name? Once we identify the location of those parts, we can use the OCR engine to extract Text.

I have a solution where I used OCR + some regex to identify the right text.

As an expected result, I need to get Patient Name, Date of Birth from it.

1 Answers

Luca Vavassori On

Have a look at this article: https://www.pyimagesearch.com/2018/08/20/opencv-text-detection-east-text-detector/

You can detect text and then perform OCR on the image crop containing it.