ABBYY FlexiCapture SDK

ABBYY FlexiCapture SDK includes the latest version of ABBYY OCR/ICR technologies. It provides the best recognition quality for more than 209 OCR languages, including Arabic, Chinese, Japanese, and Korean. This simplifies localization of application development and enables to meet global customers’ needs more easily.

ABBYY FlexiCapture SDK provides intelligent data extraction for certain fields. The technology looks for the fields on the document and analyzes the areas around them. To enhance the results, developers can use built-in field extraction training to more accurately define the position of fields and recurring groups on a page of a structured or a semi-structured document. The training starts right when an end-user starts working with documents.

FlexiCapture SDK allows developers to implement training for field extraction and classification technology to improve accuracy of field extraction and classification.

The technology learns based on feedback provided from end users. This feedback may be collected in any third-party component for verification, be implemented by the developer, and contain information on correct field regions (to train the field extraction technology) and on correct image classes (to train the classification technology).

Now developers can improve images, including images captured by mobile devices, both manually and automatically. The feature allows developers to work with the complex backgrounds of documents like IDs, insurance policies, etc. For instance, the following image corrections can be applied:

• Automatic correction of resolution
• Overwriting resolution of the image to a specified resolution
• Correction of the image skew angle based on black separators, black squares, or text
• Conversion of the image to black and white
• Rotation, inversion, and mirroring of the image
• Reduction of ISO noise
• Removing of color marks and garbage
• Automatic crop of the image

It is also possible to apply a pre-configured image enhancement profile to images.

Advanced document classification helps sort incoming documents into categories using a pre-configured classifier available via API. The classification technologies are highly flexible and can be used to distinguish:

• Document types like invoices, bank statements, tax forms, claims, etc.
• Document variations within a document type, such as invoices from different vendors
• Document boundaries, such as the first and subsequent pages of documents of each type

The classification technology uses two approaches:

• Image classification takes advantage of convolutional neural networks and distinguishes between documents that are visually different.
• Text classification relies on statistical and semantic text analysis and requires OCR.

Using both of these approaches enables the most reliable classification results. FlexiCapture SDK provides the API to train a classifier from scratch and uses the trained classifier inside the application to speed up document processing.

Invoice processing automation is one of the most popular scenarios for implementation into third-party systems. ABBYY FlexiCapture SDK provides an API for out-of-the-box processing of invoices, as well as purchase orders and receipts. It replaces labor-intensive data input tasks with transparent, manageable, efficient, and automated data capture based on smart document analysis and character recognition technologies. It includes pre-defined settings, validation rules, and advanced database look-up capabilities specifically tailored for processing financial documents.

The technology can identify data fields on the vast majority of invoices, purchase orders, and receipts, enabling users to start processing right away. Intelligent learning-by-feedback technology refines the extraction of data from the documents with non-standard layouts.

FlexiCapture SDK supplies ready-to-use invoice, PO, and receipt projects for many countries. The invoice project may be created either via graphical user interface in FlexiCapture Developer’s Package or directly via API in FlexiCapture SDK.

The full list of supported countries can be found here ›

The newest release of FlexiCapture SDK with Natural Language Processing (NLP) extends its capture capabilities to now include unstructured documents such as contracts, leases, articles, and agreements. With NLP technology, you can automate more business processes that require manual data entry and quickly make important document data available in your business applications to save time, money, and labor. The quality of data extraction continuously improves through additional training of NLP models by the user feedback, reducing the cost and time spent on verification.

Increase the efficiency of document processing with barcode recognition. ABBYY FlexiCapture SDK leverages powerful barcode recognition technology that can read wide range of barcode types. Now Neural Networks based, it allows extracting barcodes from various documents and is able to detect objects of arbitrary shape at a wide range of aspect ratios.

FlexiCapture SDK provides both options for creating projects: using the graphical user interface in the FlexiCapture Developer’s Package or via API. This is important in the development of document management systems, where developers may need to modify the metadata, which is extracted and available for each document right from their solution. Business process outsourcers who have many FlexiCapture projects may need to tune a specific field’s setting automatically, without needing to go through each project.

FlexiCapture SDK provides API interfaces for data verification, performed by comparing recognition results with the image of any document part. The results of the verification can be used for teaching the technology to improve accuracy of extraction and classification.

To simplify initial setup of data capture process, together with FlexiCapture 12 SDK, the distributive developer receives the FlexiCapture 12 Developer’s Package. It includes the FlexiCapture Project Setup Station and the FlexiLayout Studio. The Developer’s Package may be used to set up FlexiCapture projects for further usage in FlexiCapture SDK. All FlexiCapture projects created in FlexiCapture Developer’s Package and shipped with FlexiCapture SDK may be used in FlexiCapture SDK and vice versa. 

ABBYY FlexiCapture SDK provides tools for extracting data both from fixed forms and from semi-structured and unstructured documents. All required data fields can be quickly located and extracted in an appropriate format.

ABBYY FlexiCapture SDK processes office documents in all popular formats. All documents can be processed in the same flow with no need to separate image formats from text formats. SDK supports text documents in DOC, DOCX, RTF, HTML, TXT, and ODT formats; XLS, XLSX, and ODS tables; and presentations in PPT, PPTX, and ODP.

ABBYY FlexiCapture SDK supports all modern environments like virtual machines, containers (including Docker Containers), and public cloud services, which allow for the building of highly scalable solutions such as cloud-based web services powered by container orchestrators.

Manage the size and quality of output files to achieve maximum efficiency in data storage policy. Export options include the ability to save all relevant data into backend applications and to convert analyzed documents into searchable PDF or PDF/A formats for archiving purposes. Data capture technologies export data to structured formats (like XML or CSV) or other storage locations. For some formats, the images of the original documents are also saved to make them easy find and compare if needed. Leverage a secure connection to transfer files using SFTP server, and save exporting data as a JSON file.