Comparing Documents
| Feature | Description |
|---|---|
| New “Compare Documents” Module | For quick verification of the document’s integrity, the new “Compare Documents” Module in ABBYY FineReader Engine enables detecting content differences in two versions of the same document. |
| Comparison of bilingual documents | The new option of the “Compare Documents” Module provides the ability to automatically detect the bilingual nature of such a document and its complex layout and to compare each column (and thus each language version) separately. |
Input of Office formats in Linux and Windows
| Feature | Description |
|---|---|
| Processing of Office documents | In addition to a broad set of image formats, FineReader Engine can now process input documents that are created in one of these Office document formats:
Microsoft Office or LibreOffice must be installed on the system. |
MRZ Capture
| Feature | Description |
|---|---|
| Data capture from a Machine-Readable Zone (MRZ) | The new feature allows automatic data extraction from a machine-readable zone (MRZ) in ID documents and allows faster entering and verification of personal data during customer onboarding or verification processes. |
Improved Japanese OCR
| Feature | Description |
|---|---|
| Leading recognition accuracy | With the new version of ABBYY Fine Reader Engine, Japanese OCR has seen some major improvements, bringing recognition accuracy to a new level previously unattainable for most solutions. |
Improved Arabic OCR
| Feature | Description |
|---|---|
| End-to-end recognition for Arabic on poor images | Arabic OCR on low-quality images where general technology provides low confident results with a lot of errors. |
Improved Korean OCR
| Feature | Description |
|---|---|
| Deep learning language model for Korean | A trained model for Korean language selects the best word recognition variant from recognition hypotheses or even generates new one based on a recognition context (preceding and following words). |
New neural network-based OCR technologies
| Feature | Description |
|---|---|
| Improvements in OCR technologies | With the help of neural network approaches in OCR technologies, ABBYY FineReader Engine was enhanced with handwritten and handprinted Latin symbols processing:
|
| Machine learning barcode recognition technology | The neural network architecture introduces a new model of barcode recognition performing detection of the approximate region of a barcode, its classification, and obtaining the output represented as a region with the most likely type of barcode. |
| New recognition mode | The new Accurate mode allows you to get the maximum quality of the output document, assuming a reasonable slowdown in the recognition speed. This mode is best suited for low-quality or photo-generated invoices, contracts, receipts, and ID cards. |
OCR quality improvements for text near stamps and signatures
| Feature | Description |
|---|---|
| Detecting text near stamps and signatures | Whenever an agreement contains stamps or signatures, the text nearby is recognized separately from them, thus improving the quality of the processed documents. |
New licensing options
| Feature | Description |
|---|---|
| Online License usage as Network and Standalone | The Developer’s Help for FineReader Engine 12 has been extended by additional information about different possibilities to license the SDK, describing the individual types of licensing options in an easy-to-understand comparison table. |
| Using grace periods | With the new option, customers can use the ABBYY FineReader Engine license for some time after the expiration date, thereby enlarging the license validity period. |
ICR and OMR technologies in Linux and macOS version
| Feature | Description |
|---|---|
| Handwritten text and checkmark recognition | With ABBYY FineReader Engine 12, you may recognize handwritten, handprinted characters and the checkmarks of various types. ICR and OMR technologies are implemented to extract the data from the handwritten documents and develop new data extraction solutions. |
Ability to run Engine in cloud environments
| Feature | Description |
|---|---|
| New deployment options | New licensing type allows deployment in Virtual and Cloud environments, allowing you to offer a broader spectrum of solutions. The licensing mechanism requires internet connection and supports proxy servers. Note: Applicable to FineReader Engine for Linux and Windows. |
.NET Core wrapper in FRE for Windows
| Feature | Description |
|---|---|
| New development framework | To increase the efficiency of development teams using containers and other native environments for the popular way of software development and deployment, ABBYY FineReader Engine now offers a pre-built .NET Core 6 wrapper. |
New libraries in ABBYY FineReader Engine
| Feature | Description |
|---|---|
| NeoML library usage | NeoML is an open-source end-to-end machine learning framework that allows you to build, train, and deploy Machine Learning models. This framework is used by engineers for computer vision and natural language processing tasks, including image preprocessing, classification, document layout analysis, OCR, and data extraction from structured and unstructured documents. |
| Embedded PDFium for processing PDFs | PDFium is a cross-platform native library conforming to PDF standards and controlling all operations related to PDF, including processing, parsing, rendering, and obtaining the output. |
Enhanced Document Classification
| Feature | Description |
|---|---|
| Document Classification using NLP and Machine Learning | With ABBYY FineReader Engine 12, incoming documents can be automatically sorted into different categories. Machine learning, OCR and natural language processing technologies are employed to train the image-based and text-based classifiers on representative documents. The received information is then used during classification step. |
| Text-based classifier: advanced security of training data | To train and optimize the text-based classifier, documents representing each document category must be imported. In order to protect data contained in these documents, implemented hashing algorithms avoid the possibility to recover information from the sample documents. |
| Enhanced Classification Demo Sample | ABBYY FineReader Engine is able to process PDFs, scanned or photographed document images as well as documents in Office formats. To reflect this capability in the classification process, the provided pre-compiled Demo Sample for classification was enhanced and allows now to import Office documents in addition to PDFs and image formats. |
Code sample for command-line interface (CLI)
| Feature | Description |
|---|---|
| Ready-to-use code sample | With this code sample, developers can efficiently utilize ABBYY FineReader Engine libraries and integrate document processing capabilities in command-line-based applications. |
Implementation of PDF meta-data extractor
| Feature | Description |
|---|---|
| Digitally-born PDF file processing | AuxInfo is a supplementary object of PDFium providing meta-data information from a PDF file. ABBYY R&D PDFTools team implemented its own AuxInfo object working with PDFium. |
Improved PDF processing
| Feature | Description |
|---|---|
| Improvements for PDF with “mixed” contents | ABBYY FineReader Engine provides new capabilities for processing the PDF documents containing both image-only and digitally-born pages:
|
| Using additional content in PDF | To ensure more flexible forming the PDF contents, ABBYY FineReader Engine offers the new options:
|
| Feature | Description |
|---|---|
| Farsi OCR | ABBYY FineReader Engine features updated and improved Farsi recognition options, opening up the door for more effective work with documents from Iran, Afghanistan and many other countries of the Middle East. |
| Georgian OCR | The Georgian language was added as new OCR language. |
| OCR for simple mathematical formulas | Extracting characters of simple mathematical formulas allows better recognition of scientific documents containing simple single-line mathematical formulas inside the text. |
| Technical preview for Burmese OCR | Burmese OCR was added as a technical preview to highlight future capabilities. |
| Special languages for Arabic and Japanese dates capture | FineReader Engine supports special languages for field recognition in FineReader Engine for Windows. The new version adds improved date recognition in Arabic and Japanese. |
| Technical preview for Bangla OCR | Bangla OCR was added for a technical preview to demonstrate potential functionality. |
Improved document layout recreation
| Feature | Description |
|---|---|
| Improved table reconstruction | With ABBYY FineReader Engine 12, extracted tables from documents keep their formatting better than ever. |
| Detection and recreation of balanced columns | Whenever a document contains balanced columns of text (e.g., contracts, scientific papers, articles, etc.), now the initial structure stays intact, thus simplifying document processing. |
| New “single-column” document model | The main improvements of the new algorithm are in the detection and analysis of tables and charts. |
| Enhanced table structure analysis | With the improved mechanism of document conversion, ABBYY FineReader Engine can detect tables with columns of numbers in the “Accounting” format. |
Internal process optimization for faster processing
| Feature | Description |
|---|---|
| New scheme of the ILayout object iteration | A new scheme that speeds up the iteration of the ILayout object obtained after processing the document outside the main process. Note: Applicable to FineReader Engine for Linux and Windows. |
New scanning options in FRE for Windows
| Feature | Description |
|---|---|
| More scanning capabilities | ABBYY FineReader Engine 12 has lots of device-based scanning features:
|
Online documentation
| Feature | Description |
|---|---|
| Documentation available online | In addition to the built-in documentation, you can now use the online version providing “just in time” information about the features and capabilities of ABBYY FineReader Engine. |
Latest .NET Framework versions in FRE for Windows
| Feature | Description |
|---|---|
| .Net COM Interop wrappers support | The distributive now includes .Net COM Interop wrappers for the following .Net Framework versions:
|
New export formats
| Feature | Description |
|---|---|
| JSON | JSON (JavaScript Object Notation) is as an open-standard, language-independent file format to transmit data objects consisting of attribute–value pairs and array data types. FineReader Engine now supports exporting OCR results in JSON format. |
| New ALTO versions | ALTO (Analyzed Layout and Text Object) is an XML Schema that details technical metadata to describe the layout and content of physical text resources, such as the pages of a book or newspaper. The latest versions of this schema (4.0, 4.1, 4.2) are supported in FineReader Engine 12. |
| PDF/A-2b and PDF/A-3b | PDF/A is an ISO-standardized version of the Portable Document Format (PDF), specialized for use in archiving and the long-term preservation of electronic documents. Now, FineReader Engine supports all PDF/A conformance levels. |
| Doclang | Doclang is an AI‑optimized XML format designed to represent document structure in a semantically meaningful way for modern AI pipelines, including LLM ingestion, document RAG systems, and structured document processing workflows. Available as of v. 12.8.0. |