Content Accuracy % with PDF Extract API

[Bhupathi26112575jjjn]

Hi Team,

Can you confirm the content accuracy to extract the text from PDF using API service? I've verified the below link and there is no details for the content accuracy.

https://developer.adobe.com/document-services/docs/overview/pdf-extract-api/howtos/extract-api/

Regards,

Bhupathi L

[Joel Geraci]

How do you define accuracy?

[Bhupathi26112575jjjn]

This measures the percentage of characters that were correctly converted, including letters, numbers, and special characters from PDF.

[Joel Geraci]

For native PDF (non-scanned) it's 100%. Unlike most of the structured text extraction competitors, we don't convert the PDF to an image and then OCR the text, we read it directly out of the PDF.

For scanned PDF, our accuracy is on par with other OCR tools at near 100% but very dependent on image quality.

[Bhupathi26112575jjjn]

Thanks for the details. Will review few samples and get back to you