Version 2.0.0

New Features

・Added support for recognizing the table regions with AI when converting PDF improved the accuracy of table recognition.

・Added support for accurate file conversion with OCR improved the conversion effect of scanned PDF files, and optimized the image loss problems when converting files.

・Added support for recognizing the image with OCR when converting PDF to Excel, and enhanced the picture rendering effect in the Excel table.

・Added support for recognizing and extracting the highlight, underline, squiggly, and strikeout in PDF files, and keeping these annotation features after converting them to Word, Excel, PPT, and HTML.

・Added support for converting PDF to Word or PPT with hyperlinks and keeping the links working properly.

・Added support for extracting text and text coordinates from PDF files, and output structured data as JSON files.

・Added support for extracting tables from PDF files and output structured data as JSON files.

Issues Addressed

・Optimized the way to call the C/C++ interface of the underlying library, with better operation consistency, more flexible error handling, and dynamic parameter passing, as well as higher performance and efficiency.

・Upgraded text detection model and text recognition model to improve OCR accuracy and precision.

・Optimized the line detection effect of the table recognition function for a low-resolution image.
・Optimized the process logic of table recognition. If it fails to recognize the standard tables, the non-standard table algorithm will be used for recognition again to improve the table recognition success rate.

Contact us

Email: support@compdf.com
Find All Release Here: https://www.compdf.com/api/changelog-compdfkit-api