ICR sample code - Sample code or SDK help for ICR

Hi,

I am writing some application using LeadTools. My code works good for typed documents but very poor performance on using hand written documents, even on cases where handwriting is pretty good (as attached). Do you guys have any advice on proving these results.

I am using this function for my ocr conversion " L_OcrPage_FromBitmap(hEngine, &hPage, &LeadBitmap, L_OcrBitmapSharingMode_AutoFree, NULL, NULL);" Let me know if you need more details.




I really appreciate your help.

TBone,

Working with ICR can always be somewhat tricky because it definitely is much harder for a machine to recognize handwritten text rather than machine printed text. After looking at the image you provided, I was able to improve the recognition slightly by running two simple image processing commands, the DespeckleCommand and the MinimumCommand.

• The DespeckleCommand removes small speckled dots from an image. Typically, this command is used to clean up scanned images like the one you shared.
  
  
• The MinimumCommand dilates dark objects like the handwritten text by a specified amount. This is used to make the writing more bold to help fill in pixels that may be missing.

However, when it comes to wanting the most accurate results when using ICR, it's best to use the block format as seen below so that each character has its own designated area. This will ensure that none of the letters in the word are overlapping which can cause confusion come time to recognize the file.



If you have any other questions regarding this, please email us at support@leadtools.com

Thanks Nick. If I understand you right, I need to do some prep work on those images and use the same OCR functions for conversion. Aka there is no seperate function/module I can use for ICR, is that correct?

Do you happen to have some sample code/VS solution using these functions I can use to get a feel for this type of process?

Thanks again and I really appreciate your assistance.

Regards,
Ramesh (aka Tbone) :)

Ramesh,

These are just basic Image Processing commands that you can use to help clear up your image and improve recognition during the ICR process.

You can find an example of how to use these commands in another forum post that can be found here:
https://www.leadtools.co...ve-your-Images#post42927

thanks guys. I will check it out.