OCR opens opportunities for the Vision Impaired

Written By: Peter Abrahams
Content Copyright © 2007 Bloor. All Rights Reserved.

There are many hurdles in an office for people with vision impairments. Probably the single biggest is the amount of information that still arrives in printed hard-copy: orders, instruction manuals, reference books, magazines, legislation, requests for information, etc., etc. Not having access to this printed information will limit the jobs that a vision-impaired employee can perform and may reduce the chances of finding a job at all.

It is now theoretically possible that all this information could be delivered electronically in formats suitable for screen readers or magnifiers. However, making a request for an electronic version can be seen as a barrier and for legal or copyright reasons may sometimes be denied.

Optical Character Recognition (OCR) has been around for many years and with its ability to convert printed documents into electronic documents would appear to remove the barrier. The problem has been that the processes involved in scanning and distributing electronic copies have created their own barriers. It has been too difficult, too slow and too expensive to implement.

Nuance’s announcement of OmniPage 16 includes a series of functions that lower or remove the barrier to using OCR to support people with vision impairments, they include improved batch automation, one click automation, form data retrieval, and digital camera text capture.

Batch Automation

When a batch of paper arrives a cover page can be added that defines how it should be processed and where it will be saved. The batches, including the cover pages, are then fed into the OCR machine without requiring any further human intervention. This reduces the barrier to delivering the information to a vision-impaired user.

One click automation

Clicking one button on the scanner or one button in an OmniPage window can start a workflow. For example putting a page on a local scanner and then pressing scan can start a process that scans the document, converts it into a Word document, stores it in the user’s local file system, opens the document and finally starts reading the document to the vision-impaired user. This enables a vision-impaired user to deal with printed documents completely independently; this is particularly important if the document is confidential or personal.

Form data retrieval

PDF fillable forms enable a user to complete a form electronically and then either print the form and then mail it, or send back the form electronically.

On receipt of an electronic filled form OmniPage can extract the user data and store it as a comma-separated values (CSV) file. If the filled form is on paper then OmniPage can scan the page and extract the user data. Again the vision-impaired user has access to the required information.

Digital Camera capture

Not all documents can be scanned. They may be: too big, attached to a wall or in a location without a scanner. The advent of the digital camera means that all these documents can be captured. OmniPage has been extended to extract text from digital camera images. The problem with these images is they are taken in three dimensions, unlike a scanner that only has two dimensions. The OCR technology has to be able to recognise text that is not aligned, or has been taken at an angle or from a curved surface (such as the page in a book). OmniPage can now compensate for these distortions and successfully extract the required text.

Performance and accuracy

For OCR to be usable by vision-impaired users:

  • It must perform well to minimise any delays but also to reduce the cost that might otherwise be a continuing barrier.
  • The accuracy must be high otherwise the user will still be dependent on sighted colleagues to fully understand the documents.

OmniPage 16 has made significant improvements in performance and accuracy that enable it to be used by vision-impaired users it nearly all situations.

OCR as the norm

All the functions described above have been primarily developed to enhance normal business processes, they just happen to be hugely beneficial to vision-impaired users.

Most users, both those with a disability and those without, will be more productive processing the electronic version of a document, rather than the bulky hardcopy, as they will not have the problems of distribution, storage and sharing.

In many situations today input documents are scanned as a matter of course because it improves the productivity of the process. As a by-product it creates a level playing field for people with vision-impairments to enter the job market.

An enterprises use of OCR can, not only improve general productivity, but also enable them to employ more people with disabilities. Employing people with disabilities has major social, moral, legal and financial benefits.

Companies should investigate and invest in the use of OCR.