Tessaract Ocr read Specific location pdf

[KhalidAnsari]

Hi,

I need to read specific location from .pdf that has digital signature on it. I have gone through forum related this, I have seen screen capture and Tessaract example. I don't known how to get exact location. 

Please help me out with this.

Thanks,

Khalid Ansari

[FrancescoDiMuro]

@KhalidAnsari
Digging a little bit on the Internet, you can read data from a PDF file without using Tesseract OCR.
You can directly use VBA (translate it to AutoIt)

[KhalidAnsari]

Hi @FrancescoDiMuro

Thanks for reply. I have converted this .net but i need to read digital signature that is in image format. So i though i should use Tessaract and screen capture.

Any other help will be appreciated.

Thanks,

Khalid

[FrancescoDiMuro]

@KhalidAnsari
Then you can take a look at this UDF (or something similiar available on the Forum)

[KhalidAnsari]

Hi @FrancescoDiMuro

Thanks for reply. I can able to find the position through screen capture UDF. Out i am getting in Junk character. 

Actual text = "Login For Chat"

Output = "Ln-gin fur Chat |"

Am I missing any character recognition for English?

Following are the tessdata file detail

Thanks,

Khalid 

 

Edited January 9, 2019 by KhalidAnsari

[FrancescoDiMuro]

@KhalidAnsari
How do you call the Tesseract OCR from your script?

[KhalidAnsari]

Hi 

Thanks for quick reply.

I am calling it this way 

Local $TESS_EXE="C:\Program Files (x86)\Tesseract-OCR\tesseract.exe"

 

[FrancescoDiMuro]

@KhalidAnsari
Could you please post the entire code you are using?

[KhalidAnsari]

@FrancescoDiMuro

 

Attached is my sample file which i am using.OcrScreenCapture.au3

Thanks,

Khalid

[FrancescoDiMuro]

@KhalidAnsari
As mentioned in the Tesseract OCR Wiki, try to use the parameter -l:

$TESS_PARAMS= $TIF_FILENAME & " output -l eng"

 

Edited January 9, 2019 by FrancescoDiMuro

[KhalidAnsari]

@FrancescoDiMuro

Thanks for quick reply.

I m making change then I post my output.

Thanks,

Khalid

 

[KhalidAnsari]

Hi @FrancescoDiMuro

Same out put after changing that line. 

$TESS_PARAMS= $TIF_FILENAME & " output -l eng" . Below is my tiff image.

thanks

Khalidtest.tif

Edited January 9, 2019 by KhalidAnsari

[FrancescoDiMuro]

@KhalidAnsari
Sorry if I ask, but could you provide the .pdf or the download link?

[KhalidAnsari]

Hi @FrancescoDiMuro

Thanks for reply. I really appreciate ur effort.

Following this I am doing

1. Read the Application Title name. Based on that I will use winWait till that application window load.

2. Then i need to pass data from application.

3. Then i need to open pdf file and again i need to read OCr detail for specific location of pdf .

Point 1 and 3 are for OCR.

Attaching sample pdf which don't have digital signature.

Thanks,

Khalid

invoice1.pdf

Edited January 9, 2019 by KhalidAnsari

[KhalidAnsari]

Hi

Any suggestions or help

Thanks

[FrancescoDiMuro]

@KhalidAnsari

A little research on the Forum

[KhalidAnsari]

Hi @FrancescoDiMuro

 I have search the forum and able to read the pdf and tiff file. I am facing difficulty while read this type of hand written scanned pdf data. Out put result is in txt file. Please review it

 

Result.txt

Edited January 22, 2019 by KhalidAnsari

[KhalidAnsari]

Any help please.

Thanks,

Khalid

[KhalidAnsari]

Hi @FrancescoDiMuro

Any suggestion will be appreciated

Thanks,

 

[FrancescoDiMuro]

@KhalidAnsari

Tesseract OCR is a free OCR software, so, it has its limits.

If you want more from you're OCR software, you may spend some money with a kore reliable (and non-free) software, which could help you with this kind of recognition