最佳 C# OCR 函式庫：IronOCR vs Tesseract vs Azure AI OCR vs Aspose.OCR

using IronOcr;

var ocr = new IronTesseract();

// Configure advanced preprocessing
ocr.Configuration.ReadBarCodes = true;
ocr.Configuration.WhiteListCharacters = "ABCDEFGHIJKLMNOPQRSTUVWXYZ0123456789";

// Apply preprocessing filters
using var input = new OcrInput(@"invoice.jpg");
input.Deskew();
input.DeNoise();
input.EnhanceResolution(300);

// Perform OCR with confidence scores
var result = ocr.Read(input);
Console.WriteLine($"Text: {result.Text}");
Console.WriteLine($"Confidence: {result.Confidence}%");

// Export as searchable PDF
result.SaveAsSearchablePdf("searchable-invoice.pdf");

Imports IronOcr

Dim ocr As New IronTesseract()

' Configure advanced preprocessing
ocr.Configuration.ReadBarCodes = True
ocr.Configuration.WhiteListCharacters = "ABCDEFGHIJKLMNOPQRSTUVWXYZ0123456789"

' Apply preprocessing filters
Using input As New OcrInput("invoice.jpg")
    input.Deskew()
    input.DeNoise()
    input.EnhanceResolution(300)

    ' Perform OCR with confidence scores
    Dim result = ocr.Read(input)
    Console.WriteLine($"Text: {result.Text}")
    Console.WriteLine($"Confidence: {result.Confidence}%")

    ' Export as searchable PDF
    result.SaveAsSearchablePdf("searchable-invoice.pdf")
End Using

using Tesseract;

// Basic Tesseract usage - requires manual preprocessing
using (var engine = new TesseractEngine(@"./tessdata", "eng", EngineMode.Default))
{
    // Set page segmentation mode
    engine.DefaultPageSegMode = PageSegMode.Auto;

    using (var img = Pix.LoadFromFile(@"document.png"))
    {
        using (var page = engine.Process(img))
        {
            var text = page.GetText();
            var confidence = page.GetMeanConfidence();

            Console.WriteLine($"Mean confidence: {confidence}");
            Console.WriteLine($"Text: {text}");

            // 手動的 hOCR export
            var hocr = page.GetHOCRText(1);
        }
    }
}

Imports Tesseract

' Basic Tesseract usage - requires manual preprocessing
Using engine As New TesseractEngine("./tessdata", "eng", EngineMode.Default)
    ' Set page segmentation mode
    engine.DefaultPageSegMode = PageSegMode.Auto

    Using img As Pix = Pix.LoadFromFile("document.png")
        Using page As Page = engine.Process(img)
            Dim text As String = page.GetText()
            Dim confidence As Single = page.GetMeanConfidence()

            Console.WriteLine($"Mean confidence: {confidence}")
            Console.WriteLine($"Text: {text}")

            ' 手動的 hOCR export
            Dim hocr As String = page.GetHOCRText(1)
        End Using
    End Using
End Using

using Aspose.OCR;

// Configure Aspose.OCR with preprocessing
var api = new AsposeOcr();
var settings = new RecognitionSettings
{
    AutoSkew = true,
    AutoDenoising = true,
    Language = Language.Eng,
    RecognitionAreas = new List<Rectangle>()
};

// Process document with structure preservation
var result = api.RecognizeImage("invoice.png", settings);

// Access structured results
foreach (var line in result.RecognitionAreasText)
{
    Console.WriteLine($"Text: {line.Text}");
    Console.WriteLine($"Confidence: {line.Confidence}");
}

// Export as searchable PDF
api.SaveMultipageDocument("output.pdf", SaveFormat.Pdf, result);

Imports Aspose.OCR

' Configure Aspose.OCR with preprocessing
Dim api As New AsposeOcr()
Dim settings As New RecognitionSettings With {
    .AutoSkew = True,
    .AutoDenoising = True,
    .Language = Language.Eng,
    .RecognitionAreas = New List(Of Rectangle)()
}

' Process document with structure preservation
Dim result = api.RecognizeImage("invoice.png", settings)

' Access structured results
For Each line In result.RecognitionAreasText
    Console.WriteLine($"Text: {line.Text}")
    Console.WriteLine($"Confidence: {line.Confidence}")
Next

' Export as searchable PDF
api.SaveMultipageDocument("output.pdf", SaveFormat.Pdf, result)

using Azure.AI.FormRecognizer.DocumentAnalysis;
using Azure;

// Configure Azure AI client
var credential = new AzureKeyCredential("your-api-key");
var client = new DocumentAnalysisClient(
    new Uri("your-endpoint"), credential);

// Analyze document with OCR
var operation = await client.AnalyzeDocumentFromUriAsync(
    WaitUntil.Completed, 
    "prebuilt-read", 
    new Uri("___PROTECTED_URL_143___"));

var result = operation.Value;

// Process structured results
foreach (var page in result.Pages)
{
    Console.WriteLine($"Page {page.PageNumber}");

    foreach (var line in page.Lines)
    {
        Console.WriteLine($"Text: {line.Content}");
        Console.WriteLine($"Confidence: {line.Confidence}");

        // Access bounding box coordinates
        var boundingBox = line.BoundingPolygon;
    }
}

// Extract tables if present
foreach (var table in result.Tables)
{
    Console.WriteLine($"Table with {table.RowCount} rows");
}

Imports Azure.AI.FormRecognizer.DocumentAnalysis
Imports Azure

' Configure Azure AI client
Dim credential As New AzureKeyCredential("your-api-key")
Dim client As New DocumentAnalysisClient(New Uri("your-endpoint"), credential)

' Analyze document with OCR
Dim operation = Await client.AnalyzeDocumentFromUriAsync(WaitUntil.Completed, "prebuilt-read", New Uri("___PROTECTED_URL_143___"))

Dim result = operation.Value

' Process structured results
For Each page In result.Pages
    Console.WriteLine($"Page {page.PageNumber}")

    For Each line In page.Lines
        Console.WriteLine($"Text: {line.Content}")
        Console.WriteLine($"Confidence: {line.Confidence}")

        ' Access bounding box coordinates
        Dim boundingBox = line.BoundingPolygon
    Next
Next

' Extract tables if present
For Each table In result.Tables
    Console.WriteLine($"Table with {table.RowCount} rows")
Next