在 C# 中快速配置 OCR 以获得最佳性能

Curtis Chau

Updated: 2026年5月9日

IronOCR的快速配置通过使用EnglishFast语言模式并禁用条形码读取等不必要的功能实现了高达17%的OCR处理速度提升，而不损失准确性。这种优化非常适合时间紧迫的大批量处理。

IronOCR 开箱即可有效运行。当速度优先于绝对准确性时，IronOCR 可提供快速配置。此设置可大幅提高扫描性能，而对准确性的影响却很小，因此比标准的 OCR 配置要快得多。

本文演示了如何设置快速配置，并比较了快速配置和标准 IronOCR 配置的基准结果。无论您是在处理扫描文档、PDF，还是在处理图像，这些优化都能显著提高应用程序的性能。

如何设置 OCR 快速配置

安装OCR库并使用NuGet设置IronOcr
初始化 OCR 引擎
将Language 设置为EnglishFast
将ReadBarCodes属性设置为false
加载图像并提取文本

快速开始：在 C# 中配置快速 OCR

快速配置的主要组件是Language属性。将OcrLanguage.EnglishFast优先考虑速度，而不是以小的准确度潜在损失为代价。这使得 IronOCR 的批量阅读速度大大加快，这在时间至关重要的任务关键型应用中尤其有用。

除了设置快速语言模式之外，您还可以通过禁用诸如ReadBarCodes等不必要的配置来获得进一步的速度提升。让 IronOCR 自动检测页面分割，以保持设置简单。有关更多高级配置选项，请参阅我们的 Tesseract 详细配置指南。

下面的代码示例处理了以下输入图像：

我应该使用什么输入格式？

快速配置需要哪些代码？

1Install IronOCR with NuGet Package Manager
PM > Install-Package IronOcr
```
Install-Package IronOcr
```

2复制并运行这段代码。

/* :path=/static-assets/ocr/content-code-examples/how-to/ocr-fast-configuration.cs */
using IronOcr;
using System;

var ocrTesseract = new IronTesseract();

// Fast Dictionary
ocrTesseract.Language = OcrLanguage.EnglishFast;

// Turn off unneeded options
ocrTesseract.Configuration.ReadBarCodes = false;

// Assume text is laid out neatly in an orthogonal document
ocrTesseract.Configuration.PageSegmentationMode = TesseractPageSegmentationMode.Auto;

using var ocrInput = new OcrInput();
ocrInput.LoadImage("image.png");

var ocrResult = ocrTesseract.Read(ocrInput);
Console.WriteLine(ocrResult.Text);

/* :path=/static-assets/ocr/content-code-examples/how-to/ocr-fast-configuration.cs */
using IronOcr;
using System;

var ocrTesseract = new IronTesseract();

// Fast Dictionary
ocrTesseract.Language = OcrLanguage.EnglishFast;

// Turn off unneeded options
ocrTesseract.Configuration.ReadBarCodes = false;

// Assume text is laid out neatly in an orthogonal document
ocrTesseract.Configuration.PageSegmentationMode = TesseractPageSegmentationMode.Auto;

using var ocrInput = new OcrInput();
ocrInput.LoadImage("image.png");

var ocrResult = ocrTesseract.Read(ocrInput);
Console.WriteLine(ocrResult.Text);

3部署到您的生产环境中进行测试

通过免费试用立即在您的项目中开始使用IronOCR

我可以期待什么结果？

这是从上方提取的文本输出。 OCR 引擎可以准确捕捉文学文本，同时保持原始格式和结构。快速配置为像本例这样清晰、高对比度的文本提供了出色的效果。

快速配置与标准配置相比有何优势？

为了展示真实世界的影响，我们对标准配置和快速配置的性能进行了基准测试。我们使用一组 10 个样本图片（每个图片包含几个段落）来比较使用快速配置的性能和可视化权衡。

对于标准配置，我们以默认设置初始化IronTesseract，而不应用任何速度导向的属性。这种基准方法类似于我们的性能跟踪指南，它展示了如何实时监控 OCR 操作。

以下是我们用于运行测试的输入示例。这些图像代表了您在处理多页文档或批处理操作时可能遇到的典型文档场景。

如何运行基准测试？

using IronOcr;
using System;
using System.Diagnostics;
using System.IO;

// --- Tesseract Engine Setup ---
var ocrTesseract = new IronTesseract();
ocrTesseract.Language = OcrLanguage.EnglishFast;
ocrTesseract.Configuration.ReadBarCodes = false;
ocrTesseract.Configuration.PageSegmentationMode = TesseractPageSegmentationMode.Auto;

// --- 1. Define folder and get files ---
string folderPath = @"images"; // IMPORTANT: Set this to your image directory
string filePattern = "*.png";    // Change to "*.jpg", "*.bmp", etc. as needed
string outputFilePath = "ocr_results.txt"; // The new results file

// Get all image files in the directory
var imageFiles = Directory.GetFiles(folderPath, filePattern);

Console.WriteLine($"Found {imageFiles.Length} total images to process...");
Console.WriteLine($"Results will be written to: {outputFilePath}");

// --- 2. Start timer and process images, writing to file ---
// Open the output file *before* the loop for efficiency
using (StreamWriter writer = new StreamWriter(outputFilePath))
{
    var stopwatch = Stopwatch.StartNew();

    foreach (var file in imageFiles)
    {
        string fileName = Path.GetFileName(file);

        using var ocrInput = new OcrInput();
        ocrInput.LoadImage(file);

        var ocrResult = ocrTesseract.Read(ocrInput);

        // Check if any text was actually found
        if (!string.IsNullOrEmpty(ocrResult.Text))
        {
            // Write to Console
            Console.WriteLine($"--- Text found in: {fileName} ---");
            Console.WriteLine(ocrResult.Text.Trim());
            Console.WriteLine("------------------------------------------");

            // Write to File
            writer.WriteLine($"--- Text found in: {fileName} ---");
            writer.WriteLine(ocrResult.Text.Trim());
            writer.WriteLine("------------------------------------------");
            writer.WriteLine(); // Add a blank line for readability
        }
        else
        {
            // Write to Console
            Console.WriteLine($"No text found in: {fileName}");

            // Write to File
            writer.WriteLine($"No text found in: {fileName}");
            writer.WriteLine();
        }
    }

    stopwatch.Stop();

    // --- 3. Print and write final benchmark summary ---
    string lineSeparator = "\n========================================";
    string title = "Batch OCR Processing Complete";
    string summary = $"Fast configuration took {stopwatch.Elapsed.TotalSeconds:F2} seconds";

    // Write summary to Console
    Console.WriteLine(lineSeparator);
    Console.WriteLine(title);
    Console.WriteLine("========================================");
    Console.WriteLine(summary);

    // Write summary to File
    writer.WriteLine(lineSeparator);
    writer.WriteLine(title);
    writer.WriteLine("========================================");
    writer.WriteLine(summary);

    if (imageFiles.Length > 0)
    {
        string avgTime = $"Average time per image: {(stopwatch.Elapsed.TotalSeconds / (double)imageFiles.Length):F3} seconds";
        Console.WriteLine(avgTime);
        writer.WriteLine(avgTime);
    }
}

Console.WriteLine($"\nSuccessfully saved results to {outputFilePath}");

using IronOcr;
using System;
using System.Diagnostics;
using System.IO;

// --- Tesseract Engine Setup ---
var ocrTesseract = new IronTesseract();
ocrTesseract.Language = OcrLanguage.EnglishFast;
ocrTesseract.Configuration.ReadBarCodes = false;
ocrTesseract.Configuration.PageSegmentationMode = TesseractPageSegmentationMode.Auto;

// --- 1. Define folder and get files ---
string folderPath = @"images"; // IMPORTANT: Set this to your image directory
string filePattern = "*.png";    // Change to "*.jpg", "*.bmp", etc. as needed
string outputFilePath = "ocr_results.txt"; // The new results file

// Get all image files in the directory
var imageFiles = Directory.GetFiles(folderPath, filePattern);

Console.WriteLine($"Found {imageFiles.Length} total images to process...");
Console.WriteLine($"Results will be written to: {outputFilePath}");

// --- 2. Start timer and process images, writing to file ---
// Open the output file *before* the loop for efficiency
using (StreamWriter writer = new StreamWriter(outputFilePath))
{
    var stopwatch = Stopwatch.StartNew();

    foreach (var file in imageFiles)
    {
        string fileName = Path.GetFileName(file);

        using var ocrInput = new OcrInput();
        ocrInput.LoadImage(file);

        var ocrResult = ocrTesseract.Read(ocrInput);

        // Check if any text was actually found
        if (!string.IsNullOrEmpty(ocrResult.Text))
        {
            // Write to Console
            Console.WriteLine($"--- Text found in: {fileName} ---");
            Console.WriteLine(ocrResult.Text.Trim());
            Console.WriteLine("------------------------------------------");

            // Write to File
            writer.WriteLine($"--- Text found in: {fileName} ---");
            writer.WriteLine(ocrResult.Text.Trim());
            writer.WriteLine("------------------------------------------");
            writer.WriteLine(); // Add a blank line for readability
        }
        else
        {
            // Write to Console
            Console.WriteLine($"No text found in: {fileName}");

            // Write to File
            writer.WriteLine($"No text found in: {fileName}");
            writer.WriteLine();
        }
    }

    stopwatch.Stop();

    // --- 3. Print and write final benchmark summary ---
    string lineSeparator = "\n========================================";
    string title = "Batch OCR Processing Complete";
    string summary = $"Fast configuration took {stopwatch.Elapsed.TotalSeconds:F2} seconds";

    // Write summary to Console
    Console.WriteLine(lineSeparator);
    Console.WriteLine(title);
    Console.WriteLine("========================================");
    Console.WriteLine(summary);

    // Write summary to File
    writer.WriteLine(lineSeparator);
    writer.WriteLine(title);
    writer.WriteLine("========================================");
    writer.WriteLine(summary);

    if (imageFiles.Length > 0)
    {
        string avgTime = $"Average time per image: {(stopwatch.Elapsed.TotalSeconds / (double)imageFiles.Length):F3} seconds";
        Console.WriteLine(avgTime);
        writer.WriteLine(avgTime);
    }
}

Console.WriteLine($"\nSuccessfully saved results to {outputFilePath}");

Imports IronOcr
Imports System
Imports System.Diagnostics
Imports System.IO

' --- Tesseract Engine Setup ---
Dim ocrTesseract As New IronTesseract()
ocrTesseract.Language = OcrLanguage.EnglishFast
ocrTesseract.Configuration.ReadBarCodes = False
ocrTesseract.Configuration.PageSegmentationMode = TesseractPageSegmentationMode.Auto

' --- 1. Define folder and get files ---
Dim folderPath As String = "images" ' IMPORTANT: Set this to your image directory
Dim filePattern As String = "*.png" ' Change to "*.jpg", "*.bmp", etc. as needed
Dim outputFilePath As String = "ocr_results.txt" ' The new results file

' Get all image files in the directory
Dim imageFiles = Directory.GetFiles(folderPath, filePattern)

Console.WriteLine($"Found {imageFiles.Length} total images to process...")
Console.WriteLine($"Results will be written to: {outputFilePath}")

' --- 2. Start timer and process images, writing to file ---
' Open the output file *before* the loop for efficiency
Using writer As New StreamWriter(outputFilePath)
    Dim stopwatch = Stopwatch.StartNew()

    For Each file In imageFiles
        Dim fileName As String = Path.GetFileName(file)

        Using ocrInput As New OcrInput()
            ocrInput.LoadImage(file)

            Dim ocrResult = ocrTesseract.Read(ocrInput)

            ' Check if any text was actually found
            If Not String.IsNullOrEmpty(ocrResult.Text) Then
                ' Write to Console
                Console.WriteLine($"--- Text found in: {fileName} ---")
                Console.WriteLine(ocrResult.Text.Trim())
                Console.WriteLine("------------------------------------------")

                ' Write to File
                writer.WriteLine($"--- Text found in: {fileName} ---")
                writer.WriteLine(ocrResult.Text.Trim())
                writer.WriteLine("------------------------------------------")
                writer.WriteLine() ' Add a blank line for readability
            Else
                ' Write to Console
                Console.WriteLine($"No text found in: {fileName}")

                ' Write to File
                writer.WriteLine($"No text found in: {fileName}")
                writer.WriteLine()
            End If
        End Using
    Next

    stopwatch.Stop()

    ' --- 3. Print and write final benchmark summary ---
    Dim lineSeparator As String = vbLf & "========================================"
    Dim title As String = "Batch OCR Processing Complete"
    Dim summary As String = $"Fast configuration took {stopwatch.Elapsed.TotalSeconds:F2} seconds"

    ' Write summary to Console
    Console.WriteLine(lineSeparator)
    Console.WriteLine(title)
    Console.WriteLine("========================================")
    Console.WriteLine(summary)

    ' Write summary to File
    writer.WriteLine(lineSeparator)
    writer.WriteLine(title)
    writer.WriteLine("========================================")
    writer.WriteLine(summary)

    If imageFiles.Length > 0 Then
        Dim avgTime As String = $"Average time per image: {(stopwatch.Elapsed.TotalSeconds / CDbl(imageFiles.Length)):F3} seconds"
        Console.WriteLine(avgTime)
        writer.WriteLine(avgTime)
    End If
End Using

Console.WriteLine(vbLf & $"Successfully saved results to {outputFilePath}")

Imports IronOcr
Imports System
Imports System.Diagnostics
Imports System.IO

' --- Tesseract Engine Setup ---
Dim ocrTesseract As New IronTesseract()
ocrTesseract.Language = OcrLanguage.EnglishFast
ocrTesseract.Configuration.ReadBarCodes = False
ocrTesseract.Configuration.PageSegmentationMode = TesseractPageSegmentationMode.Auto

' --- 1. Define folder and get files ---
Dim folderPath As String = "images" ' IMPORTANT: Set this to your image directory
Dim filePattern As String = "*.png" ' Change to "*.jpg", "*.bmp", etc. as needed
Dim outputFilePath As String = "ocr_results.txt" ' The new results file

' Get all image files in the directory
Dim imageFiles = Directory.GetFiles(folderPath, filePattern)

Console.WriteLine($"Found {imageFiles.Length} total images to process...")
Console.WriteLine($"Results will be written to: {outputFilePath}")

' --- 2. Start timer and process images, writing to file ---
' Open the output file *before* the loop for efficiency
Using writer As New StreamWriter(outputFilePath)
    Dim stopwatch = Stopwatch.StartNew()

    For Each file In imageFiles
        Dim fileName As String = Path.GetFileName(file)

        Using ocrInput As New OcrInput()
            ocrInput.LoadImage(file)

            Dim ocrResult = ocrTesseract.Read(ocrInput)

            ' Check if any text was actually found
            If Not String.IsNullOrEmpty(ocrResult.Text) Then
                ' Write to Console
                Console.WriteLine($"--- Text found in: {fileName} ---")
                Console.WriteLine(ocrResult.Text.Trim())
                Console.WriteLine("------------------------------------------")

                ' Write to File
                writer.WriteLine($"--- Text found in: {fileName} ---")
                writer.WriteLine(ocrResult.Text.Trim())
                writer.WriteLine("------------------------------------------")
                writer.WriteLine() ' Add a blank line for readability
            Else
                ' Write to Console
                Console.WriteLine($"No text found in: {fileName}")

                ' Write to File
                writer.WriteLine($"No text found in: {fileName}")
                writer.WriteLine()
            End If
        End Using
    Next

    stopwatch.Stop()

    ' --- 3. Print and write final benchmark summary ---
    Dim lineSeparator As String = vbLf & "========================================"
    Dim title As String = "Batch OCR Processing Complete"
    Dim summary As String = $"Fast configuration took {stopwatch.Elapsed.TotalSeconds:F2} seconds"

    ' Write summary to Console
    Console.WriteLine(lineSeparator)
    Console.WriteLine(title)
    Console.WriteLine("========================================")
    Console.WriteLine(summary)

    ' Write summary to File
    writer.WriteLine(lineSeparator)
    writer.WriteLine(title)
    writer.WriteLine("========================================")
    writer.WriteLine(summary)

    If imageFiles.Length > 0 Then
        Dim avgTime As String = $"Average time per image: {(stopwatch.Elapsed.TotalSeconds / CDbl(imageFiles.Length)):F3} seconds"
        Console.WriteLine(avgTime)
        writer.WriteLine(avgTime)
    End If
End Using

Console.WriteLine(vbLf & $"Successfully saved results to {outputFilePath}")

该基准代码演示了几个重要概念：

1.批量处理：代码在单个操作中处理多个图像，类似于我们的多线程 OCR 示例，该示例展示了如何利用并行处理来提高速度。

性能测量：使用Stopwatch类提供精确到毫秒的正确时间测量，这是比较不同配置的关键。

3.结果记录：控制台和文件输出可确保您稍后分析结果并验证不同配置之间的准确性差异。

我可以期待哪些性能提升？

模式	总时间	平均耗时/图像	与标准相比，时间增益	与标准相比，精度提升
标准	10.40秒	1.040秒	Baseline	Baseline
快速	8.60秒	0.860秒	+17.31%（速度更快）	+0%（相同）

标准配置和快速配置之间的基准比较显示，快速配置具有显著的性能优势。通过将标准模式作为基准（总耗时 10.40 秒），快速配置仅用 8.60 秒就完成了同一批 10 张图片的翻译。这意味着时间上节省了17.31%。最重要的是，速度的提高并没有影响质量；两种模式的准确性完全相同，两种配置产生的文本输出也相同。

要验证结果，您可以下载快速文本输出和标准文本输出。

何时应该使用快速配置？

快速配置对以下方面尤其有益

大量文档处理，数千页文档需要快速处理
实时应用程序，响应时间至关重要
需要保持响应式用户体验的 Web 应用程序
批量处理系统，运行时间紧迫

对于涉及多种语言、低质量扫描或牌照或护照等专业文档类型的更复杂的情况，您可能需要使用标准配置来确保最大的准确性。

IronOCR 使配置之间的切换变得简单--只需更改几个属性，您的应用程序就能适应不同的性能要求，而无需对代码进行重大修改。

常见问题解答

快速 OCR 配置比标准设置快多少？

与标准 OCR 设置相比，IronOCR 的快速配置可使处理速度提高 17%，而对准确性的影响却微乎其微。这种性能提升是通过 EnglishFast 语言模式和禁用不必要的功能实现的。

实现快速 OCR 处理的主要设置是什么？

在 IronOCR 中实现快速配置的主要组件是将语言属性设置为 OcrLanguage.EnglishFast。这将速度放在首位，而不是在准确性方面付出较小的潜在代价，因此非常适合批量处理和时间紧迫的应用程序。

除了使用 EnglishFast 模式外，如何进一步优化 OCR 速度？

您可以通过禁用 IronOCR 中不必要的功能来提高速度，例如，如果不需要条形码检测，可将 ReadBarCodes 设置为 false。此外，还可以使用 TesseractPageSegmentationMode.Auto 让 IronOCR 自动检测页面分割。

何时应使用快速 OCR 配置而不是标准设置？

IronOCR 中的快速 OCR 配置非常适合大批量处理场景，在这些场景中，时间至关重要，而在准确性方面稍作权衡是可以接受的。对于需要快速处理扫描文档、PDF 或图像的关键任务应用程序来说，它尤其有用。

快速配置是否适用于所有文档类型？

是的，IronOCR 的快速配置可有效处理各种文档类型，包括扫描文档、PDF 和图像。无论您处理的是哪种输入格式，都能获得优化优势。

使用快速 OCR 模式是否会降低准确性？

IronOCR 的快速配置可显著提高扫描性能，而对准确性的影响却微乎其微。虽然在使用 EnglishFast 模式时可能会在准确性方面付出一些潜在的代价，但对于优先考虑速度的应用程序来说，这种权衡往往是值得的。

Can IronOCR's fast configuration handle batch processing efficiently?

Yes, IronOCR's fast configuration is designed for efficient batch processing, allowing multiple images to be processed in a single operation with enhanced speed due to reduced computational load per image.

How does benchmark testing demonstrate the performance of fast configuration?

Benchmark testing with a set of sample images revealed that fast configuration can process images 17% quicker than the standard mode, demonstrating its superior speed without compromising text output accuracy.

When should the standard OCR configuration be preferred over fast configuration?

The standard OCR configuration should be used in scenarios requiring high accuracy, such as processing low-quality scans, recognizing multiple languages, or handling specialized documents like passports or license plates.

Is it easy to switch between fast and standard configurations in IronOCR?

Yes, IronOCR allows easy switching between configurations by changing a few properties, enabling swift adaptation to varying performance needs without complex code adjustments.