Full Source Code

        API Library

        Plug-in Module


        Scanners, Cameras

        Mobile Devices

        Existing Image Files


Excerpt from Document Imaging Report on February 7, 2014. Click here for the pdf version of the article.

The Avis of IP?

Dynamic Computing Solutions is a Maryland-based ISV and scanning service bureau that has been working on PerfectScan for more than a decade. “When we started out developing PerfectScan, our goal was to create something that would be as good as VRS,” said Randy McNees, founder of DCS. “Our thinking was that the market was big enough that we could be Pepsi to Kofax’s Coke.”

The road from product conception to realization was a lot longer than McNees anticipated. “We made a lot of mistakes along the way,” he admitted. “Every time we thought we had it done, it turned out we didn’t.”

Until now. A demo version of PerfectScan is now available and McNees took DIR through a fairly extensive demo himself. “One of the characteristics that differentiates PerfectScan from most IP technology is its ability to make adjustments considering the entire contents of a page,” said McNees. “Most other image processing software takes a more myopic view and focuses more on individual characters. PerfectScan considers a page the way a human eye and brain would. This enables it to make only a section lighter or darker, where needed. We call this localized dynamic Thresholding.”

McNees showed us what he meant utilizing a weigh bill with a sticky note attached to it. After initially be run through the default settings in PerfectScan, the words on the sticky notes were illegible. However, with an adjustment of a slider, the contrast on just the sticky note area was increased, enabling the words to come into view, while the contrast on the rest of the document, where the text legibility was fine originally, remained the same.

The slider was a light-dark control appearing at the top of the screen. The area of the sticky note did not need to be highlighted or circled. The software inherently knew where the contrast needed to be adjusted. This type of simplicity of interface is a key feature of PerfectScan. On the demo version, there are only three controls: the light-dark slider, three buttons to select light, dark or medium mode, and an option for selecting an “enhanced” mode.

“The application appears simple to users, but whenever an adjustment is made, behind the scenes there are multiple operations going on,” said McNees. “If a customer wanted us to, we could expose additional controls.” The enhanced processing mode offers features like “intergap fill,” which automatically completes lines on characters that may appear broken in a scanned image. “This is not character substitution and it is language independent technology. It works on handprint as well as machine print,” said McNees. “Our tests have shown that on poor quality documents, PerfectScan can improve OCR accuracy by 38% over using an OCR product’s out-of-the-box binarization algorithm.”

PerfectScan works with grayscale images, and an algorithm for converting color images to bi-tonal is currently in development. We saw it work effectively on a very low contrast image captured with a camera phone—for an application in which no outside light sources are allowed to be used. PerfectScan can also be used with microfilm images. Speed is also a focus, and the images we saw were processed at a rate of well less than one second per page.

Dynamic Computing Solutions is a relatively small operation, and McNees is currently looking for partners, potentially through OEM agreements, to help him bring PerfectScan to market.

For more information: http://perfectscan.com/