MODI OCR C# 与 IronOCR：在 C# 中选择合适的 OCR 光学字符识别库

// tessdata/eng.traineddata must exist before this line runs
// Downloaded separately: curl -L -o tessdata/eng.traineddata
//   https://github.com/tesseract-ocr/tessdata_best/raw/main/eng.traineddata
using var engine = new Engine(@"./tessdata", Language.English, EngineMode.Default);
using var image = TesseractOCR.Pix.Image.LoadFromFile("document.png");
using var page = engine.Process(image);

string text = page.Text;
float confidence = page.MeanConfidence; // Returns 0.0-1.0 float

Imports TesseractOCR

' tessdata/eng.traineddata must exist before this line runs
' Downloaded separately: curl -L -o tessdata/eng.traineddata
'   https://github.com/tesseract-ocr/tessdata_best/raw/main/eng.traineddata
Using engine As New Engine("./tessdata", Language.English, EngineMode.Default)
    Using image As Pix.Image = TesseractOCR.Pix.Image.LoadFromFile("document.png")
        Using page As Page = engine.Process(image)
            Dim text As String = page.Text
            Dim confidence As Single = page.MeanConfidence ' Returns 0.0-1.0 float
        End Using
    End Using
End Using

// From BasicOcrService in tesseractocr-basic-ocr.cs
private void ValidateTessData()
{
    if (!Directory.Exists(_tessDataPath))
    {
        throw new DirectoryNotFoundException(
            $"tessdata folder not found at: {_tessDataPath}\n" +
            "Download traineddata files from: https://github.com/tesseract-ocr/tessdata_best");
    }

    string engTrainedData = Path.Combine(_tessDataPath, "eng.traineddata");
    if (!File.Exists(engTrainedData))
    {
        throw new FileNotFoundException(
            $"eng.traineddata not found in {_tessDataPath}\n" +
            "Download from: https://github.com/tesseract-ocr/tessdata_best/raw/main/eng.traineddata");
    }
}

Private Sub ValidateTessData()
    If Not Directory.Exists(_tessDataPath) Then
        Throw New DirectoryNotFoundException(
            $"tessdata folder not found at: {_tessDataPath}" & vbCrLf &
            "Download traineddata files from: https://github.com/tesseract-ocr/tessdata_best")
    End If

    Dim engTrainedData As String = Path.Combine(_tessDataPath, "eng.traineddata")
    If Not File.Exists(engTrainedData) Then
        Throw New FileNotFoundException(
            $"eng.traineddata not found in {_tessDataPath}" & vbCrLf &
            "Download from: https://github.com/tesseract-ocr/tessdata_best/raw/main/eng.traineddata")
    End If
End Sub

// Multi-language requires each .traineddata file pre-downloaded
// eng.traineddata + fra.traineddata + deu.traineddata all required
using var engine = new Engine(@"./tessdata",
    Language.English | Language.French | Language.German,
    EngineMode.Default);

// If any traineddata file is missing, this throws at construction time
using var image = TesseractOCR.Pix.Image.LoadFromFile(imagePath);
using var page = engine.Process(image);

return page.Text;

Imports TesseractOCR

' Multi-language requires each .traineddata file pre-downloaded
' eng.traineddata + fra.traineddata + deu.traineddata all required
Using engine As New Engine("./tessdata", Language.English Or Language.French Or Language.German, EngineMode.Default)

    ' If any traineddata file is missing, this throws at construction time
    Using image As TesseractOCR.Pix.Image = TesseractOCR.Pix.Image.LoadFromFile(imagePath)
        Using page As Page = engine.Process(image)
            Return page.Text
        End Using
    End Using
End Using

// dotnet add package IronOcr.Languages.French
// dotnet add package IronOcr.Languages.German
// 否 tessdata folder, no download scripts, no path validation

var ocr = new IronTesseract();
ocr.Language = OcrLanguage.French;
ocr.AddSecondaryLanguage(OcrLanguage.German);

using var input = new OcrInput();
input.LoadImage(imagePath);
var result = ocr.Read(input);

return result.Text;

Imports IronOcr

Dim ocr As New IronTesseract()
ocr.Language = OcrLanguage.French
ocr.AddSecondaryLanguage(OcrLanguage.German)

Using input As New OcrInput()
    input.LoadImage(imagePath)
    Dim result = ocr.Read(input)
    Return result.Text
End Using

// Requires: dotnet add package SixLabors.ImageSharp
// Manual preprocessing — each parameter requires tuning per document type

using var image = Image.Load(imagePath);

image.Mutate(x => x.Grayscale());
image.Mutate(x => x.Contrast(1.5f));         // 1.5 is a guess; tune per use case
image.Mutate(x => x.GaussianBlur(0.5f));     // Denoise with blur
image.Mutate(x => x.BinaryThreshold(0.5f)); // Threshold requires manual tuning

// Deskew is NOT in ImageSharp — requires separate Hough transform implementation
// (~50-100 additional lines)

string tempPath = Path.GetTempFileName() + ".png";
try
{
    image.Save(tempPath);

    using var engine = new Engine(@"./tessdata", Language.English);
    using var pixImage = TesseractOCR.Pix.Image.LoadFromFile(tempPath);
    using var page = engine.Process(pixImage);

    return page.Text;
}
finally
{
    File.Delete(tempPath); // Clean up temp file
}

Imports SixLabors.ImageSharp
Imports SixLabors.ImageSharp.Processing
Imports TesseractOCR
Imports System.IO

' Requires: dotnet add package SixLabors.ImageSharp
' Manual preprocessing — each parameter requires tuning per document type

Dim image As Image = Image.Load(imagePath)

image.Mutate(Sub(x) x.Grayscale())
image.Mutate(Sub(x) x.Contrast(1.5F))         ' 1.5 is a guess; tune per use case
image.Mutate(Sub(x) x.GaussianBlur(0.5F))     ' Denoise with blur
image.Mutate(Sub(x) x.BinaryThreshold(0.5F))  ' Threshold requires manual tuning

' Deskew is NOT in ImageSharp — requires separate Hough transform implementation
' (~50-100 additional lines)

Dim tempPath As String = Path.GetTempFileName() & ".png"
Try
    image.Save(tempPath)

    Using engine As New Engine("./tessdata", Language.English)
        Using pixImage As TesseractOCR.Pix.Image = TesseractOCR.Pix.Image.LoadFromFile(tempPath)
            Using page As Page = engine.Process(pixImage)
                Return page.Text
            End Using
        End Using
    End Using
Finally
    File.Delete(tempPath) ' Clean up temp file
End Try

// 否 external imaging library needed
// 否 temp files, no manual parameter tuning

using var input = new OcrInput();
input.LoadImage(imagePath);
input.Deskew();           // Automatic angle detection and correction
input.DeNoise();          // Intelligent noise removal
input.Contrast();         // 自动对比度 enhancement
input.EnhanceResolution(300); // Upscale if below 300 DPI

var result = new IronTesseract().Read(input);

return result.Text;

// Requires: dotnet add package TesseractOCR
//           dotnet add package Docnet.Core
// Note: Docnet is MIT-licensed; iTextSharp would be AGPL

using var library = DocLib.Instance;
using var docReader = library.GetDocReader(pdfPath, new PageDimensions(dpi, dpi));

int pageCount = docReader.GetPageCount();
var allText = new StringBuilder();
var tempFiles = new List<string>();

try
{
    using var engine = new Engine(_tessDataPath, Language.English, EngineMode.Default);

    for (int pageIndex = 0; pageIndex < pageCount; pageIndex++)
    {
        using var pageReader = docReader.GetPageReader(pageIndex);
        var width = pageReader.GetPageWidth();
        var height = pageReader.GetPageHeight();
        var imageBytes = pageReader.GetImage(); // BGRA bytes

        // TesseractOCR.Pix.Image requires a file path — write to temp
        string tempPath = Path.Combine(_tempDirectory, $"page_{pageIndex}_{Guid.NewGuid()}.png");
        tempFiles.Add(tempPath);
        SaveBgraAsPng(imageBytes, width, height, tempPath); // ~30 lines

        using var image = TesseractOCR.Pix.Image.LoadFromFile(tempPath);
        using var page = engine.Process(image);

        allText.AppendLine($"--- Page {pageIndex + 1} ---");
        allText.AppendLine(page.Text);
    }
}
finally
{
    foreach (var tempFile in tempFiles)
    {
        try { File.Delete(tempFile); } catch { }
    }
}

Imports Docnet.Core
Imports TesseractOCR
Imports System.IO
Imports System.Text

' Requires: dotnet add package TesseractOCR
'           dotnet add package Docnet.Core
' Note: Docnet is MIT-licensed; iTextSharp would be AGPL

Dim library = DocLib.Instance
Dim docReader = library.GetDocReader(pdfPath, New PageDimensions(dpi, dpi))

Dim pageCount As Integer = docReader.GetPageCount()
Dim allText As New StringBuilder()
Dim tempFiles As New List(Of String)()

Try
    Using engine As New Engine(_tessDataPath, Language.English, EngineMode.Default)
        For pageIndex As Integer = 0 To pageCount - 1
            Using pageReader = docReader.GetPageReader(pageIndex)
                Dim width = pageReader.GetPageWidth()
                Dim height = pageReader.GetPageHeight()
                Dim imageBytes = pageReader.GetImage() ' BGRA bytes

                ' TesseractOCR.Pix.Image requires a file path — write to temp
                Dim tempPath As String = Path.Combine(_tempDirectory, $"page_{pageIndex}_{Guid.NewGuid()}.png")
                tempFiles.Add(tempPath)
                SaveBgraAsPng(imageBytes, width, height, tempPath) ' ~30 lines

                Using image = TesseractOCR.Pix.Image.LoadFromFile(tempPath)
                    Using page = engine.Process(image)
                        allText.AppendLine($"--- Page {pageIndex + 1} ---")
                        allText.AppendLine(page.Text)
                    End Using
                End Using
            End Using
        Next
    End Using
Finally
    For Each tempFile In tempFiles
        Try
            File.Delete(tempFile)
        Catch
        End Try
    Next
End Try

// 满的 PDF — native, no external library
var ocr = new IronTesseract();
using var input = new OcrInput();
input.LoadPdf(pdfPath);
var result = ocr.Read(input);
string text = result.Text;

// 受密码保护的PDF — built-in, one parameter
using var encryptedInput = new OcrInput();
encryptedInput.LoadPdf("encrypted.pdf", Password: "secret");
var encryptedResult = ocr.Read(encryptedInput);

// Specific page range — no manual loop required
using var pageInput = new OcrInput();
pageInput.LoadPdfPages(pdfPath, startPage: 1, endPage: 5);
var pageResult = ocr.Read(pageInput);

Imports IronOcr

Dim ocr As New IronTesseract()

' 满的 PDF — native, no external library
Using input As New OcrInput()
    input.LoadPdf(pdfPath)
    Dim result = ocr.Read(input)
    Dim text As String = result.Text
End Using

' 受密码保护的PDF — built-in, one parameter
Using encryptedInput As New OcrInput()
    encryptedInput.LoadPdf("encrypted.pdf", Password:="secret")
    Dim encryptedResult = ocr.Read(encryptedInput)
End Using

' Specific page range — no manual loop required
Using pageInput As New OcrInput()
    pageInput.LoadPdfPages(pdfPath, startPage:=1, endPage:=5)
    Dim pageResult = ocr.Read(pageInput)
End Using

// WARNING: Engine is NOT thread-safe — must create per thread
// Memory: _maxDegreeOfParallelism * engine footprint (~40-100MB each)

var results = new ConcurrentDictionary<string, string>();

Parallel.ForEach(
    imagePaths,
    new ParallelOptions { MaxDegreeOfParallelism = 4 },
    imagePath =>
    {
        // Per-thread engine — required, expensive (~500ms init, ~50MB memory)
        using var engine = new Engine(@"./tessdata", Language.English, EngineMode.Default);
        using var image = TesseractOCR.Pix.Image.LoadFromFile(imagePath);
        using var page = engine.Process(image);

        results[imagePath] = page.Text;
    });

Imports System.Collections.Concurrent
Imports System.Threading.Tasks

' WARNING: Engine is NOT thread-safe — must create per thread
' Memory: _maxDegreeOfParallelism * engine footprint (~40-100MB each)

Dim results As New ConcurrentDictionary(Of String, String)()

Parallel.ForEach(
    imagePaths,
    New ParallelOptions With {.MaxDegreeOfParallelism = 4},
    Sub(imagePath)
        ' Per-thread engine — required, expensive (~500ms init, ~50MB memory)
        Using engine As New Engine("./tessdata", Language.English, EngineMode.Default)
            Using image As TesseractOCR.Pix.Image = TesseractOCR.Pix.Image.LoadFromFile(imagePath)
                Using page As Page = engine.Process(image)
                    results(imagePath) = page.Text
                End Using
            End Using
        End Using
    End Sub)

// Single instance — thread-safe, no per-thread duplication
var ocr = new IronTesseract();

var results = new ConcurrentDictionary<string, string>();

Parallel.ForEach(imagePaths, imagePath =>
{
    using var input = new OcrInput(imagePath);
    results[imagePath] = ocr.Read(input).Text;
});

Imports System.Collections.Concurrent
Imports System.Threading.Tasks

' Single instance — thread-safe, no per-thread duplication
Dim ocr As New IronTesseract()

Dim results As New ConcurrentDictionary(Of String, String)()

Parallel.ForEach(imagePaths, Sub(imagePath)
    Using input As New OcrInput(imagePath)
        results(imagePath) = ocr.Read(input).Text
    End Using
End Sub)

// Before (TesseractOCR)
using var engine = new Engine(@"./tessdata", Language.English, EngineMode.Default);
using var image = TesseractOCR.Pix.Image.LoadFromFile("document.png");
using var page = engine.Process(image);
float confidence = page.MeanConfidence; // 0.0 to 1.0

// After (IronOCR)
var ocr = new IronTesseract();
using var input = new OcrInput("document.png");
var result = ocr.Read(input);
double confidence = result.Confidence; // 0 to 100

Imports TesseractOCR
Imports IronOcr

' Before (TesseractOCR)
Using engine As New Engine("./tessdata", Language.English, EngineMode.Default)
    Using image As TesseractOCR.Pix.Image = TesseractOCR.Pix.Image.LoadFromFile("document.png")
        Using page As Page = engine.Process(image)
            Dim confidence As Single = page.MeanConfidence ' 0.0 to 1.0
        End Using
    End Using
End Using

' After (IronOCR)
Dim ocr As New IronTesseract()
Using input As New OcrInput("document.png")
    Dim result As OcrResult = ocr.Read(input)
    Dim confidence As Double = result.Confidence ' 0 to 100
End Using