OmniPage Capture SDK 19 adds enhanced form processing technologies

In the ongoing trend to digitize paper, OCR technology is critical to increasing the effectiveness and efficiency of business processes and workflows. Nuance's new Capture SDK, provides the industry's most accurate OCR and contains poweful new features that support new content workflows and formats.
By

It is widely understood that high accuracy is the paramount goal of OCR technology. Higher accuracy means higher automation level, lower manual verification, and therefore lower costs and higher return on investment. Nuance Document Imaging, with the combination of several pioneering OCR companies, has been dedicated for decades to improving the OCR accuracy and continues to do so. Our world leading OmniPage Capture SDK provides customers with the most accurate OCR on the market to power their products for their end users.

But the market requirements for document digitization and data capturing do not stop at the OCR accuracy. Document imaging technologies need to address more and more complex processes as businesses evolve. Nuance has a keen eye on these processes and leads on all fronts in the development of new technologies that address these challenges. For example:

  • In the last few years, Nuance developed a Cloud service and provided Imaging and OCR functionality for any/all device that supports HTTPS access.
  • We’ve architected it to meet the demands of numerous mobile OS platforms to enable customers to build out their OCR solutions for mobile devices.
  • As many now use smartphone cameras as a substitute for business scanners, we greatly enhanced our image processing technologies to improve the OCR accuracy on photos.
  • To help users of tablet devices with the best eBook reading experience, Nuance OmniPage Capture SDK 19 supports the ePub format.
  • The SDK now outputs audios from OCR results through Nuance’s flagship TTS technology into MP3 format, enabling virtually all users to listen to documents read aloud.

One of the biggest business processes related to document imaging is form processing. No matter how much effort is put into electronic forms, paper form processing remains unavoidable and the needs for solutions to support this continues to grow. Most form processing involves structured data extraction, which means solutions not only recognize the text on the page, but also classify the data into data fields based on the form labels so the data can be used directly in a database entry, for example. One effective approach for structured data extraction is through template-based form matching. In this case, a template is created from a sample of a known form. The template contains the critical textual and/or graphical features that uniquely identify this form from other types of forms. It also contains the locations and sizes of the data fields, along with data field names. When a filled form is sent to the form processing application, the software first extracts the same textual and/or graphical features then compares it to the template. If a match is found, the software further overlays data to the filled form sample to extract the data from the specified areas and associate it with the appropriate field names. Further, the application can be implemented to automatically match the filled form image against a library of different templates, enhancing the productivity of business processes throughout an organization.

In the latest OmniPage Capture SDK 19, Nuance includes a powerful yet simple tool, Form Template Editor or FTE, for developers and end users to create the template easily. This standalone application takes a blank or filled form and creates the template automatically using the OmniPage OCR engine and the form recognition technology. It significantly simplifies template creation, making sophisticated form recognition accessible for end-users without requiring special training.

SDK image 1

The diagram below shows how easy FTE is to use. First, users open a blank or filled form and click on the auto process page button. Through OCR, it generates the template with all the data field locations along with the field labels. It also intelligently selects 4 critical locations with unique text as anchors to be used during form matching, enhancing the reliability of the matching. Second, if the OCR created template is not perfect or if the user prefers to make some changes, the graphical editing tools provided in the UI enables easy modifications. Once the user is satisfied with the result, it is saved into a template file and is added to the template library. Last, users can open a batch of test images of the same type and/or other types to test against the saved template. This allows users to test the accuracy of matching to the correct template and, for the correct template match, the accuracy of the form field data extraction. With FTE, users are able to create production quality templates from scratch in only a few minutes.

SDK image 2

Further, users can add multiple templates of different forms into the template library and test multiple image samples against the whole library of templates. FTE determine which image matches which template, enabling users to simulate their entire form processing without writing any code. It is especially handy when users start with their form processing operation on one set of form types but need to add additional types later. FTE allows them to run all their previous testing with the added new templates and test images.

For casual form processing users, FTE can be used to complete all your tasks. Users add all the form images they want to process under the Test Images tree and run each one against all the templates in the library. For the matched template, data can be exported into a CSV file on one row. For each image that matches the same template, it will add one more row to the CSV file. Once users are done with all the test samples, they will have collected data from all the form images into the CSV file.

If users need to develop their own form processing application or add the form processing functionality into an existing application, the coding simple. All that is needed are four function calls to 1) load the template library, 2) find the template match, 3) extract field data with the matched template, and 4) export the data. The diagram below shows a simple example.

SDK image 3

At Nuance, we are dedicated to providing OCR technology that enables organizations to achieve and support support business goals while lowering costs.  The Form Template Editor and the new form processing technology are examples of technologies that enable our developers to extend provide these same benefits to their customers. For more information about the Form Template Editor, the new form processing API, or OmniPage Capture SDK 19 in general, please visit http://www.nuance.com/for-business/by-product/omnipage/csdk/index.htm or request a free evaluation copy.

Sources:

Tags: , ,

The latest in tech news and insights,

delivered straight to your inbox

Thank you, your preferences have now been saved.

Jeff Segarra

About Jeff Segarra

Jeff Segarra is Senior Director of Product Management for the Document Imaging Division at Nuance Communications. He is responsible for the team that delivers the full line of scanning and business productivity products including eCopy scanning solutions for MFPs, the PDF Converter product line for creating, editing and converting PDF files, OmniPage for OCR document conversion and PaperPort for desktop document management. He has an MBA from Iona College, Hagan School of Business and has been working in the software technology field for over 20 years. Jeff is an original New Yorker and, therefore, a staunch Yankees fan – in the heart of Red Sox nation.