zonal OCR sample in VB/C#.NET needed

[djdrew_BOLT]

can anyone provide sample code/project illustrating this?

here is my current relevant code attempt

Code: Select all

 datapath = Application.StartupPath & "\ocrdats"  
        Dim region As PXO_InputField
        region.blackList = VarPtrAny(wlist)
        region.whiteList = VarPtrAny(blist)
        region.top = 110
        region.bottom = 200
        region.nPage = 1
        region.left = 106
        region.right = 209
        Dim options As new PXO_Options
        options.blackList = VarPtrAny(wlist)
        options.whiteList = VarPtrAny(blist)
        options.ImageFlags = OcrCommon.OCR_ImageProcessingFlags.OCR_Image_NoRotate '& OcrCommon.OCR_ImageProcessingFlags.OCR_Image_SuppressOutput 'SuppressOutput only saves a hidden text pdf (will combine with original image pdf). Also turned off rotate (is actually deskew) so text positions match up with image when it is merged back together in a separate function.
        options.lang = 0 '0 = defaults to english, not currently supporting other languages
        options.raster_dpi = 300 'less than 300 and OCR suffers
        options.RegionMode = OcrCommon.OCR_RegionMode.OCR_Auto
        options.DataPath = VarPtrAny(datapath)
        options.accMode = 0

        '9/21/12 allows specifying a page number to ocr. Is 0 based
        Dim pg As Integer 'pg will either be a pointer to newpagelist or 0
        pg = 0

        '9/21/12 end page number

        
        Dim textout As String = ""
        textout = VarPtrAny(textout)
        'res = OCR_MakeSearchable(doc, options, pg) 'runs the actual OCR  whole page ocr working
        res = OCR_GetField(doc, options, textout, region, 1)
        'res = OCR_GetText(doc, options, textout, pg, )
        'SysFreeString(textout)
        OCR_SaveA(doc, "OCRFinal.pdf")

as you can see i've made two/three different attempts see commented out res = lines

Any suggestions or help is appreciated Thanks

[Walter-Tracker Supp]

It is unnecessary to append '\ocrdats' to the data directory; this is hardcoded into the library, so that if you specify "c:\myproject", OCR will look for language files in "c:\myproject\ocrdats". In your case, it will be looking for "(Applicationpath)\ocrdats\ocrdats\".

I realize this is an unnecessary restriction, but since we released with this behaviour in place we haven't seen fit to remove it.

It'll be changed in the new SDK versions coming out in 2013, so that the path provided will be exactly the path to search for language files.

[djdrew_BOLT]

thanks, but either way it still doesn't work, error code is -2113732591

[Walter-Tracker Supp]

Here's a quick and dirty C# example which does include both a call to OCR_MakeSearchable() and OCR_GetFields().

I realize you posted VB code, but you mentioned both C# and VB and at the moment (under a big deadline) it is most expedient to give this particular example.

It might help if you could provide the PDF you are trying to OCR, or one that at least reprdoces the problems you are having. Is the error code returned from OCR_GetFields() or is it from OCR_MakeSearchable()? Remember coordinates are in PDF points, and specifying "1" for the flags (last argument) means the top left corner is the origin, with the positive Y axis pointing down.

[djdrew_BOLT]

your sample works, probably because you helped develop the sdk lol, but mine still does not, although it no longer returns an error code

[John - Tracker Supp]

Certainly helps to be involved at the 'coal face'

If you still have problems - perhaps strip your code down to that which is relevant and send to us - along with a specific PDF and I am sure walter can apply a little 'magic.

cheers

[djdrew_BOLT]

i still don't have a clue what i was doing wrong, but i built a new fresh project and it's all fine *shrugs* thanks everyone

[John - Tracker Supp]

Wierd - but pleased all is well