Sunday, June 11, 2006 12:02 PM
Bill Bither
DotImage Advanced Document Cleanup
In our upcoming release of DotImage 4.0 we are inculding a new add-on module called DotImage Advanced Document Cleanup. This is something that our scientists have been working on for quite some time, and will add tremendous value to our product offering and and customers who choose to license it. This module can be used to clean-up scanned documents
Here are some examples of what it can do:
Hole Punch Removal
This algorithm works with many types of hole punches, including the square spiral bound holes. It will automatically calculates the hole size, or allow the user to specify the minimum and maximum hole diameter.
Auto Invert Text
This algorithms is very useful prior to OCR. It does a great job detecting white on black regions that OCR engines will typically ignore.
Border Removal
Useful for removing borders from pages that are scanned on a flatbed scanner. Improves compression, and OCR accuracy.
Line Removal
Useful for improving OCR accuracy. This algorithm will reconnect characters that were broken from the line removal. There are 5 different line removal commands including solid line removal, dotted line removal, broken line removal, and form line removal.
Other algorithms available in this product include Auto Border Crop, Auto Negate Page, Blank Page Detection, Blob Removal, Margin Crop, Noise Removal, and Speck Removal. I'll provide more examples in future posts.
All of these algorithms and more are available in the beta that was released on Friday. You can try them out with the demo which is installed with the product (Advanced Document Cleanup Demo). Licensing will be $1999 for the SDK which includes 20 runtimes. Additional runtimes cost $99 per machine or $999 per server. This is in additional to DotImage Document Imaging which is a required dependency.
We have tested these algorithms on many images, but this is a beta product and we're looking for our customers to test these out on other images and report any problems. Problems can be posted in our beta forms.
Now I'm off to Tech-Ed. Hope to see some of you there!!
