PDFs with columns and PNGs with text oriented at different angles

Question

Hi,
For PDFs with columns, such as this:
![screenshot of german language pdf with columns](/media/postImages/original/IMtsWfCpfYQo6dJek-soEacw)
The Textract result reads the text left to right across the page. Are there any workarounds for this?

For PNGs with text oriented at different angles, such as this:
![screenshot of german spatial plan with text oriented at different angles](/media/postImages/original/IMQbnC_E5XSx6OT75bji0Ijw)
The Textract results picks up some information, but not all:
![screenshot of textract results from german spatial plan with text oriented at different angles ](/media/postImages/original/IMKnjfFDo7Qpah5kllRFCwOg)
Are there any workarounds for this?

Answer

For multi-column documents, I'd suggest trying getLineClustersInReadingOrder() from the JavaScript/TypeScript version of the Amazon Textract Response parser (reading-order feature doc & code here on GitHub, package on NPM). It's only an approximate/rule-based heuristic though, so might struggle a bit with the warping in your example or need tuning the parameters a little.

For images like your second sample that are a bit more visually busy and light on text, it might be worth trying whether Rekognition DetectText can perform any better?

ESPERTO

Alex_T

con risposta un anno fa

For images like your second sample that are a bit more visually busy and light on text, it might be worth trying whether [Rekognition DetectText](https://docs.aws.amazon.com/rekognition/latest/dg/text-detection.html) can perform any better?

PDFs with columns and PNGs with text oriented at different angles

Contenuto pertinente