Textract Not Read Bold Clear Computer Generated 'X' . Need Suggestions on how to tackle this.

Question

Use Case :  We are extracting data from a table to convert it into a json file

Issue : Despite of bold and clear letters , textract is unable to read ' X ' from the table elements (pls check the attachment for better reference [image as well as json , where the data is missing (For Ex. The 'X' besides Demographics is not being captured )])

Textract API Used :  response = client.analyze_document(Document={'Bytes': bytes_test}, FeatureTypes=['TABLES'])

Remedies Tried : adjusting the image threshold, de-blurring , adjusting contrast, cropping and magnifying the object in the page with no visible success
I would greatly appreciate it if you could provide guidance, suggest best practices, or help troubleshoot the issues we are experiencing. Your expertise and assistance are crucial to the success of our project.
![![Enter image description here](/media/postImages/original/IMgPKMmbQjQfqc3BMym83DVA)
Enter image description here](/media/postImages/original/IMs94V_zubTrWijb8xGWVEhA)

Answer

Thank you for using Textract and bringing this issue to our notice. With machine learning models, we cannot guarantee 100% accuracy. We are continuously improving the accuracy of our models in response to our customer feedback.

As for now, these Xs can be either returned as a [selection element](https://docs.aws.amazon.com/textract/latest/dg/how-it-works-selectables.html) or be returned as a word when they are detected. We recommend to add some post processing logic to for addressing this issue.

Textract Not Read Bold Clear Computer Generated 'X' . Need Suggestions on how to tackle this.

Relevant content