Overview of OCR Xpress for Linux

OCR Xpress for Linux is a full-page OCR engine based on a C API. Any image file in an uncompressed BMP format can be loaded and processed without any image pre-filtering or pre-processing. OCR Xpress for Linux provides the conversion of BMP formatted images into a searchable PDF documents. One or more BMP images can be built into a single PDF document. In addition, OCR Xpress for Linux provides a rich API that allows the customer to access the same internal OCR results used to generate the PDF documents.

In addition to using OCR Xpress for Linux in an end-to-end product solution for converting full page images into searchable text, there are several other uses in which customers may apply OCR Xpress for Linux functionality. OCR Xpress for Linux can also convert an image to a TXT file for archiving searchable text. By archiving the original image with the searchable text file in a database, it can later be retrieved according to the results of searches for key words or phrases in the text file.

For applications that need to access post-OCR data for processing purposes, OCR Xpress for Linux generates and maintains an internal hierarchical model of the text it finds in an image. Every character is hierarchically tied to the word, text line, text block, region, and page with which it is associated. The same is true of every word, text line, text block, region, and page of the generated document. The rich API allows the application to access this internal hierarchical model. With OCR Xpress for Linux, a form reader application can extract data from the form based on its location. The API also provides confidence levels of the text in question so that the application can make content usage decisions based on the confidence that the recognized text is correct. The complete internal hieratical model can even be copied into an application's local work space to allow the application to perform higher levels of segmentation, content association, and analysis of the text data.

The OCR Xpress for Linux toolkit provides support in the following areas:

Full page stand-alone OCR (no pre-filtering or pre-processing of the image is required) of non-compressed BMP files.
Access to hierarchically structured OCR results for advanced post-processing.
Export to searchable PDF file (single-page or multi-page).
Export to TXT file.