This example will show how to use the IOcrEngine in a multi-threaded application. It will simultaneously convert four TIF files into PDF.

For an example on how to use the IOcrEngine in a single-threaded application, refer to IOcrEngine example.

Visual Basic

Copy Code

Private Structure MyThraedData    Public ImageFileName As String    Public WaitHandle As AutoResetEvent End Structure Public Sub CreateEngineInAppDomainExample()    ' Unlock the support needed for LEADTOOLS Plus OCR engine    RasterSupport.Unlock(RasterSupportType.Document, "Replace with your own key here")    RasterSupport.Unlock(RasterSupportType.OcrPlus, "Replace with your own key here")    RasterSupport.Unlock(RasterSupportType.OcrPlusPdfLeadOutput, "Replace with your own key here")    ' The image file names we are going to OCR and convert to PDF    Dim tifFileNames() As String = _    { _       LeadtoolsExamples.Common.ImagesPath.Path + "Ocr1.tif", _       LeadtoolsExamples.Common.ImagesPath.Path + "Ocr2.tif", _       LeadtoolsExamples.Common.ImagesPath.Path + "Ocr3.tif", _       LeadtoolsExamples.Common.ImagesPath.Path + "Ocr4.tif" _    }    Dim threadCount As Integer = tifFileNames.Length    ' Create the thread    Dim threads(threadCount - 1) As Thread    Dim waitHandles(threadCount - 1) As AutoResetEvent    For i As Integer = 0 To threadCount - 1       threads(i) = New Thread(AddressOf MyThreadProc)       threads(i).Name = "OCR thread + " + i.ToString()       waitHandles(i) = New AutoResetEvent(False)    Next    Console.WriteLine("Starting the threads and waiting...")    ' Start the threads    For i As Integer = 0 To threadCount - 1       Dim threadData As New MyThraedData()       threadData.ImageFileName = tifFileNames(i)       threadData.WaitHandle = waitHandles(i)       threads(i).Start(threadData)    Next    ' Wait till all threads are done    WaitHandle.WaitAll(waitHandles)    Console.WriteLine("All threads finished") End Sub Private Sub MyThreadProc(ByVal data As Object)    ' Grab the data    Dim threadData As MyThraedData = CType(data, MyThraedData)    Dim imageFileName As String = threadData.ImageFileName    ' Show a status message    Console.WriteLine("Begin: {0}", imageFileName)    ' Create an instance of the OCR engine using the LEADTOOLS Thunk Server    Using ocrEngine As IOcrEngine = OcrEngineManager.CreateEngine(OcrEngineType.Plus, True)       ' Start the engine using default parameters       ocrEngine.Startup(Nothing, Nothing, Nothing, Nothing)       ' Get the PDf file name       Dim pdfFileName As String = Path.ChangeExtension(imageFileName, "pdf")       ' Create an OCR document       Using ocrDocument As IOcrDocument = ocrEngine.DocumentManager.CreateDocument()          ' Add a page to the document          Dim ocrPage As IOcrPage = ocrDocument.Pages.AddPage(imageFileName, Nothing)          ' Recognize the page          ' Note, Recognize can be called without calling AutoZone or manually adding zones. The engine will          ' check and automatically auto-zones the page          ocrPage.Recognize(Nothing)          ' Save the document we have as PDF          ocrDocument.Save(pdfFileName, DocumentFormat.Pdf, Nothing)       End Using       ' Shutdown the engine       ' Note: calling Dispose will also automatically shutdown the engine if it has been started       ocrEngine.Shutdown()    End Using    Console.WriteLine("End: {0}", imageFileName)    ' Singal the main thread    threadData.WaitHandle.Set() End Sub

C#

Copy Code

private struct MyThraedData  {     public string ImageFileName;     public AutoResetEvent WaitHandle;  }  public void CreateEngineInAppDomainExample()  {     // Unlock the support needed for LEADTOOLS Plus OCR engine     RasterSupport.Unlock(RasterSupportType.Document, "Replace with your own key here");     RasterSupport.Unlock(RasterSupportType.OcrPlus, "Replace with your own key here");     RasterSupport.Unlock(RasterSupportType.OcrPlusPdfLeadOutput, "Replace with your own key here");       // The image file names we are going to OCR and convert to PDF     string[] tifFileNames =     {        LeadtoolsExamples.Common.ImagesPath.Path + "Ocr1.tif",        LeadtoolsExamples.Common.ImagesPath.Path + "Ocr2.tif",        LeadtoolsExamples.Common.ImagesPath.Path + "Ocr3.tif",        LeadtoolsExamples.Common.ImagesPath.Path + "Ocr4.tif"     };       int threadCount = tifFileNames.Length;       // Create the thread     Thread[] threads = new Thread[threadCount];     AutoResetEvent[] waitHandles = new AutoResetEvent[threadCount];     for(int i = 0; i < threadCount; i++)     {        threads[i] = new Thread(new ParameterizedThreadStart(MyThreadProc));        threads[i].Name = "OCR thread + " + i.ToString();          waitHandles[i] = new AutoResetEvent(false);     }       Console.WriteLine("Starting the threads and waiting...");       // Start the threads     for(int i = 0; i < threadCount; i++)     {        MyThraedData threadData = new MyThraedData();        threadData.ImageFileName = tifFileNames[i];        threadData.WaitHandle = waitHandles[i];        threads[i].Start(threadData);     }       // Wait till all threads are done     WaitHandle.WaitAll(waitHandles);     Console.WriteLine("All threads finished");  }    private void MyThreadProc(object data)  {     // Grab the data     MyThraedData threadData = (MyThraedData)data;     string imageFileName = threadData.ImageFileName;       // Show a status message     Console.WriteLine("Begin: {0}", imageFileName);       // Create an instance of the OCR engine using the LEADTOOLS Thunk Server     using(IOcrEngine ocrEngine = OcrEngineManager.CreateEngine(OcrEngineType.Plus, true))     {        // Start the engine using default parameters        ocrEngine.Startup(null, null, null, null);          // Get the PDf file name        string pdfFileName = Path.ChangeExtension(imageFileName, "pdf");          // Create an OCR document        using(IOcrDocument ocrDocument = ocrEngine.DocumentManager.CreateDocument())        {             // Add a page to the document           IOcrPage ocrPage = ocrDocument.Pages.AddPage(imageFileName, null);             // Recognize the page           // Note, Recognize can be called without calling AutoZone or manually adding zones. The engine will           // check and automatically auto-zones the page           ocrPage.Recognize(null);             // Save the document we have as PDF           ocrDocument.Save(pdfFileName, DocumentFormat.Pdf, null);        }          // Shutdown the engine        // Note: calling Dispose will also automatically shutdown the engine if it has been started        ocrEngine.Shutdown();     }       Console.WriteLine("End: {0}", imageFileName);       // Singal the main thread     threadData.WaitHandle.Set();  }

The OcrEngineManager and its methods are your entry to using the Leadtools.Forms.Ocr class library.

This class provides the methods you need to create an object of the IOcrEngine interface. Afterwards, you can use the properties and methods of this interface to perform your OCR tasks.

Based on the engine type passed to the "CreateEngine" methods, the OcrEngineManager will load the OCR engine defined in one of the supporting assemblies and return an interface to IOcrEngine. Use this interface and its included types to start using the Leadtools.Forms.Ocr class library. For more information about the engine types, refer to OcrEngineType

The CreateEngine method lets you create an instance of IOcrEngine, loading the corresponding Leadtools.Forms.Ocr.[EngineName].dll assembly in a specific manner as follows:

• The CreateEngine method uses the .NET 2.0 Assembly.Load method to load an assembly. You cannot unload this assembly once it has been successfully loaded.
• If the useThunkServer parameter to CreateEngine is set to true, the internal engine DLL's are loaded using the LEADTOOLS Thunk Server. For more information, refer to LEADTOOLS OCR Thunk Server.

Your application requirements determine how to call "CreateEngine", as follows:

• If your application is going to use a single instance of IOcrEngine; pass false to useThunkServer. This is the most common behavior.
• If your application is going to use multiple threads for OCR, pass true to useThunkServer. The various OCR engines provided by LEADTOOLS are not thread safe. Passing true causes the LEADTOOLS Thunk Server to load the internal OCR engine in a separate process and unload it when the thread procedure terminates, ensuring thread safety and cleanup.
• If your application is server-based, pass true to useThunkServer. Server-based applications usually use a separate thread for each "connection". Creating the OCR engine in a separate process ensures proper thread/process safety and cleanup.
• IMPORTANT: The above discussion is for the LEADTOOLS Plus and Professional engines only (the LEADTOOLS Advantage engine supports multi-threading). To ensure thread safety and cleanup, use the LEADTOOLS Thunk Server to load the internal OCR engine in a separate process and unload it when the thread procedure terminates.
• If your application is server-based, pass true to useThunkServer. Server-based applications usuallly use a separate thread for each "connection". Creating the OCR engine in a separate process ensures proper thread/process safety and cleanup.
• IMPORTANT: The above discussion is for the LEADTOOLS Plus and Professional engines only. The LEADTOOLS Advantage engine supports multi-threading and x64 platform natively and does not require the use of a "thunk server".

Either way, you use the LEADTOOLS OCR class library various interfaces, classes and structures in the same manner. Your application code should not be affected by the way you created the engine.

System.Object
   Leadtools.Forms.Ocr.OcrEngineManager

Target Platforms: Microsoft .NET Framework 3.0, Windows XP, Windows Server 2003 family, Windows Server 2008 family

Reference