Click on the logo to view a PDF description of this product.

Among the potential use of a color scanner in document imaging is in removing forms background colors to enhance document interpretation.

 

The first commercial (high volume) use of drop out background colors was in Optical Mark Recognition (OMR). OMR documents are printed with special inks that absorb infrared light.  
   
When the documents are illuminated by an infrared light source, only the color black and pencil marks are reflected. This method is still used today in most dedicated OMR scanners.

 

 
In document imaging, this technique is impractical because users do not want to have the expense of printing specialized documents and so they are often printed with ordinary ink. Also, in many instances, the document needs to be archived with as many details as possible.

With today's black-and-white document scanners, colors can be dropped out optically by using a light source containing a specific color or, for scanners using a white light source, by putting a colored optical filter in front of the CCD sensor or lens.

Although this approach seems practical for a number of applications, it has drawbacks.

First, once you select a specific colored lamp or filter, you are confined to this color scheme and the color range associated with the filter or lamp.

     
Furthermore, if the form happens to be completed with ink that is the same color as the light source, then this valuable information is lost as well. The figure to the right illustrates this unfortunate situation ( ! ).

Recovering that information is time consuming because the document has to be physically located, moved to the edit area and its content manually keyed-in.

Designing a failsafe system for these situations is very challenging and can be very expensive.



Help is on the way. High-speed color scanners are becoming more and more affordable. With these scanners, color dropout can be performed either by electronic means inside the scanner or by software in the controlling computer. Using a color scanner is flexible since one can change the dropout color scheme simply by providing a different electronic/software filter instead of physically changing the light source. Another advantage of this method is that the number of colors to be removed is not limited to the color of the light source or filter.

Removing colors inside the scanner, while providing flexibility, does not help the situation where information is written in one of the dropped out colors. The black-and-white image presented to the application will not contain this information. Again, the "offending" document will have to be physically located and manually entered. In fact, with this scheme, the black-and-white output from the scanner is equivalent to optical dropout - the only significant advantage being that multiple color schemes can be used.

   

Because Colerase performs color dropout in software, it offers the flexibility of saving a color backup copy of the document. The user application can use the jpeg file instead of physically locating offending documents and key the data from a screen.

And Colerase is fast. On a PII 600 Mhz PC, Colerase will process over 150 letter sized documents per minute.

With Colerase, selecting colors for dropout is an interactive process. The user selects an area (1) and adds all the colors inside that area to the processing file (2). This process is repeated until all colors and shades have been identified. With every change to the processing filter, the user monitors the effect (3).

 

   
Dunord offers Colerase either as a complete application or as an API for integrators. As an application, it is available in three forms.

  • The entry level version supports batch file input only and is limited to 20 images per minute.

    Suggested retail price: 995 US$

  • The intermediate version supports TWAIN compliant scanners up to 20 pages per minute and batch file input with up to 40 images per minute.

    Suggested retail price: 1,995 US$

  • The full version supports Dunord compliant scanners, TWAIN scanners and batch file input with no limit to the number of images processed per minute.

    Suggested retail price: 3,995 US$

Colerase is also available as an API. Integrators can use this versatile API to create complete custom applications. Contact us about pricing for the Colerase API and runtime license.


You may download an evaluation copy of ColErase.

The files are available either as one file or as 4 diskette sized files.

Once dowloaded, unzip the files to a temporary directory and run setup.

  • Colerase as one file (12 540 Mb)

  •    
  • Colindex help file (3881 Kb)This file is included in the large download.

  • Last modified: October 5, 2000 / ecolerasdoc
    © COPYRIGHT, 2000 Dunord Technologies inc.