OCR with visual foxpro

[ravicoder]

Hi all

I need some guidance with using OCR with VFP9. I have a requirement for development of a data entry & reporting application where data is captured from a pre-printed form and displayed on screen. Currently the data capture is done manually by data entry operators. Goal is to automate the typing work for errorless data and efficiency.

Would appreciate any help with some resources or sample codes or some such.

Thanks
Ravi

 

[tomk3]

Hi,
see my answer some minutes ago in https://www.tek-tips.com/threads/pl...-image-in-visual-foxpro.1831418/#post-7578227

 

[Scott24x7]

I have used a combination of EZTWAIN (for getting images from scanners/cameras) and TOCR which have been excellent products, and reasonably easy to integrate. Recently I have been toying with OpenAI and image processing, but it seems to be early days, though they are advancing quickly with it.
In any case, you still need a device/method to get an image scan, and for that EZTWAIN is the way to go. One-time license if you want to buy the full version and I think they have a "free" version floating around still.

 

[Gerrit Broekhuis]

I have used EZTWAIN too, but with TESSERACT, Ghostscript and ImageMagick (all free to use except EZTWAIN) to do graphic format conversion for best OCR results (convert to TIFF). EZTWAIN has been taken over years ago, but you will find it with Google.

Another option is to use GPT-4O and use AI to get text from the TIFF. This option is not free.

Both options work fine, however I still tend to prefer TESSERACT, since the result is somewhat more predictable.

Of course you can allow your users to choose which option they want to use.

Whatever you choose, it’s a lot of work to get every step working. Think for example about documents with multiple pages and you don’t need the text from every page. I created a preview form for this, allowing to select every page required.

Regards, Gerrit