Vis2 - Image to Text OCR()

[paulpma]

Hi, Iseahound,

Speaking of accuracy, I was thinking of using blacklist and creating two different ocr()'s. One for alpha and one for numeric. Is there a blacklist function like whitelist?

Thank you.

[iseahound]

Update: There is some heavy development going on at the moment. Going forward, there will be less of a dependence on Tesseract. Just a reminder that this project is not dead.

[robodesign]

Very good to hear this!

Good luck and thank you for developing it further.

[hasantr]

Hi, I like this work very much. Thank you.
My English is not good.

I don't understand how to run this function from a macro I created.

I want to send the coordinates and copy the text. But how.

Thanks.

[adegard]

Update: There is some heavy development going on at the moment. Going forward, there will be less of a dependence on Tesseract. Just a reminder that this project is not dead.

I tried the tool and it gives excellent result. I hope new development could reduce package size (24 Mo)and dependences
Thank you so much!

[Frosti]

Hello Iseahound, I like your tesseract wrapper and I'am watching your progress since month's. I'am using your library for some experiments with pdf files.
I missed a solution to use my own leptonica and tesseract directory's. More than about 3 times I changed your Vis2.ahk file at this position to the path's I use.

Code: Select all

      class Tesseract {

         static leptonica := A_ScriptDir "\bin\leptonica_util\leptonica_util.exe"
         static tesseract := A_ScriptDir "\bin\tesseract\tesseract.exe"
         static tessdata_best := A_ScriptDir "\bin\tesseract\tessdata_best"
         static tessdata_fast := A_ScriptDir "\bin\tesseract\tessdata_fast"

Today I just wanted to write a function in the hope that you find this useful. The function will write's the path of tesseract, tessdata and leptonica_util to an .ini file.
If someone want to use tesseract from your repository the function will write and read relative path's. So you can leave a ready made .ini file in your \lib directory of your repository.
Is this file is deleted or some key are deleted or missed, my function will ask for it.

This is an ini-file with relative paths:

Code: Select all

[leptonica]
exePath=\bin\leptonica_util\leptonica_util.exe
[tesseract]
exePath=\bin\tesseract\tesseract.exe
data_best=\bin\tesseract\tessdata_best
tessdata_fast=\bin\tesseract\tessdata_fast

Part1:

Code: Select all

      class Tesseract {

         static leptonica     		:= Vis2.stdlib.paths("leptonica")
         static tesseract     		:= Vis2.stdlib.paths("tesseract")
         static tessdata_best	    := Vis2.stdlib.paths("tessdata_best")
         static tessdata_fast		:= Vis2.stdlib.paths("tessdata_fast")

           ...

Part2:

Code: Select all

   class stdlib {

           ...
   
      ; for user customable paths to leptonica and tesseract
       paths(toWhat){
                      
			static lastdir
			
            If Instr(toWhat, "leptonica")
            {
					 IniRead, leptonica, % A_ScriptDir "\lib\Vis2.ini", leptonica, exePath
					 If ( (leptonica="ERROR") or !InStr(leptonica, "leptonica_util.exe") )
					 {
						   Vis2.Graphics.Subtitle.Render("You have not specified a path to leptonica_util.exe.`nPlease select the file now!", "t7000 xCenter y5% p1.35% c88EAB6 r8", "s2.23% cBlack")
						   again:=1
						   While (again=1)
						   {
								 FileSelectFile, leptonica,, C:, Please select leptonica.exe, *.exe
								 If InStr(leptonica, "leptonica_util.exe")
										  again=0
						   }
						   If Instr(leptonica, A_ScriptDir)
						   		 leptonica:= StrReplace(leptonica, A_ScriptDir)
						   IniWrite, % leptonica, % A_ScriptDir "\lib\Vis2.ini", leptonica, exePath
						   If ErrorLevel
								 throw Exception("Can not write to ini file!",, A_ScriptDir "\lib\Vis2.ini")
					 }

					 lastDir:= Vis2.stdlib.GetParentDir(leptonica)
					 return SubStr(leptonica, 1, 1) = "\" ? A_ScriptDir . leptonica : leptonica
            }
			
            If Instr(toWhat, "tesseract")
            {
					 IniRead, tesseract, % A_ScriptDir "\lib\Vis2.ini", tesseract, exePath
					 If ( (tesseract="ERROR") or !InStr(tesseract, "tesseract.exe") )
					 {
						   Vis2.Graphics.Subtitle.Render("You have not specified a path to tesseract.exe.`nPlease select the file now!", "t7000 xCenter y5% p1.35% c88EAB6 r8", "s2.23% cBlack")
						   again:=1
						   While (again=1)
						   {
								 FileSelectFile, tesseract,, % lastDir, Please select tesseract.exe, *.exe
								 If InStr(tesseract, "tesseract.exe")
												again=0
						   }
						   If Instr(tesseract, A_ScriptDir)
						   		 tesseract:= StrReplace(tesseract, A_ScriptDir)
						   IniWrite, % tesseract, % A_ScriptDir "\lib\Vis2.ini", tesseract, exePath
							If ErrorLevel
								 throw Exception("Can not write to ini file!",, A_ScriptDir "\lib\Vis2.ini")
					 }
					 lastDir:= Vis2.stdlib.GetParentDir(tesseract)
					 return SubStr(tesseract, 1, 1) = "\" ? A_ScriptDir . tesseract : tesseract
            }
			
            If Instr(toWhat, "tessdata_best")
            {
					 IniRead, tessdata_best, % A_ScriptDir "\lib\Vis2.ini", tesseract, data_best
					 If (tessdata_best="ERROR") 
					 {
						   Vis2.Graphics.Subtitle.Render("You have not specified a path to tessdata_best.`nPlease select the path now!", "t7000 xCenter y5% p1.35% c88EAB6 r8", "s2.23% cBlack")
						   again:=1
						   While (again=1)
						   {
								 FileSelectFolder, tessdata_best, % lastDir, 2, Please select path to tessdata_best
								 again:=ErrorLevel
						   }
						   If Instr(tessdata_best, A_ScriptDir)
						   		 tessdata_best:= StrReplace(tessdata_best, A_ScriptDir)
						   IniWrite, % tessdata_best, % A_ScriptDir "\lib\Vis2.ini", tesseract, data_best
						   If ErrorLevel
							  throw Exception("Can not write to ini file!",, A_ScriptDir "\lib\Vis2.ini")
					 }
					 lastDir:= Vis2.stdlib.GetParentDir(tessdata_best)
					 return SubStr(tessdata_best, 1, 1) = "\" ? A_ScriptDir . tessdata_best : tessdata_best
            }
			
            If Instr(toWhat, "tessdata_fast") 
            {
					 IniRead, tessdata_fast, % A_ScriptDir "\lib\Vis2.ini", tesseract, tessdata_fast
					 If (tessdata_fast="ERROR")
					 {
						   Vis2.Graphics.Subtitle.Render("You have not specified a path to tessdata_fast.`nPlease select the path now!", "t7000 xCenter y5% p1.35% c88EAB6 r8", "s2.23% cBlack")
						   again:=1
						   While (again=1)
						   {
								 FileSelectFolder, tessdata_fast, % lastDir, 2, Please select path to tessdata_fast
								 again:=ErrorLevel
						   }
						   If Instr(tessdata_fast, A_ScriptDir)
						   		 tessdata_fast:= StrReplace(tessdata_fast, A_ScriptDir)
						   IniWrite, % tessdata_fast, % A_ScriptDir "\lib\Vis2.ini",  tesseract, tessdata_fast
						   If ErrorLevel
							  throw Exception("Can not write to ini file!",, A_ScriptDir "\lib\Vis2.ini")
					 }
					 return SubStr(tessdata_fast, 1, 1) = "\" ? A_ScriptDir . tessdata_fast : tessdata_fast
            }
           
      }
   
	  ; small RegEx function to get parent dir from a given string
	  GetParentDir(Dir){                                                                                                    	
			Return RegExReplace(Dir, "\\[^\\]+$")
	  }
   
   }

   class Text {

[blue83]

Hi,

First, Iseahound this is a beautiful tool, thank you for that.

Second, is there a way to find coordinate position of string that I am searching for.

Usually there is only one unique string, but somethimes there are few of them on same page.

[blue83]

And one more question is how to increase limit of characters to be able read the whole picture from once?

For now I only get part of the page as a result.