Deep Scan for Improved PDF Data Extraction

Enhancement Description: We've introduced a "Perform Deep Scan" option to address issues with complex and poorly formatted PDFs that can affect data extraction. This feature significantly improves the accuracy of extracting data from PDFs, particularly those with hidden layers or non-standard formats.

• Deep Scan Option: Users can now request a deep scan when the reader makes mistakes, such as missing line items or dropping lines in the middle of a table. This option allows the reader to clone the PDF data pixel by pixel and reconvert it for better formatting.

• User-Directed Deep Scan: Simply select Perform Deep Scan, specify the start and end pages of the invoice, and select Deep Scan. The system will create a cloned copy of the invoice, reformat it, and rerun the extraction process.

• Improved Accuracy: This process is designed to fix extraction issues or significantly improve the accuracy of data extraction, especially for invoices spanning multiple pages.

This enhancement is particularly valuable for users relying on line item data for tax or discount calculations, as it reduces the need for manual adjustments and improves overall data integrity. To avoid unnecessary overhead, the deep scan feature is limited to direct requests from users.

‍