AWS Textract OCR reading PDF as single line instead of preserving line breaks

Question

Hi I am new to AWS Textract.

I am using Amazon Textract to extract text from a PDF file. However, the output is not preserving the line breaks from the original PDF.

For example, in the PDF there are separate lines like:

Seller

Buyer

But in the Textract output, it is reading it as: Seller: Buyer:

Instead of separate lines, the text is concatenated into a single string.

I would like Textract to retain the line breaks and structure from the original PDF. The lines denote different sections so I need to preserve that formatting.

Is there any way to configure Textract to output multi-line strings instead of concatenating everything into a single line? Or does it require post-processing the Textract result to split it based on the line breaks?

Any suggestions on how to properly extract text from a PDF while keeping the original line structure would be appreciated.

Answer 1

You're in luck, as Textract just released a new feature which might be helpful to your use case specifically.

Amazon Textract launches Layout feature to extract paragraphs, titles, and more from documents

---