Asprise OCR SDK vs IronOCR: .NET OCR Library

The Engine Lifecycle Problem

Every Asprise OCR call requires explicit engine management. The pattern is four mandatory steps: static global setup, instance creation, engine start with a language string, and engine stop after completion. Skipping engine stop leaks native resources because the garbage collector cannot release unmanaged memory:

// Asprise: four required steps before reading a single image
Ocr.SetUp();                              // Static global init
Ocr ocr = new Ocr();
try
{
    ocr.StartEngine("eng", Ocr.SPEED_FAST);   // Allocates native engine
    string text = ocr.Recognize(
        imagePath,
        Ocr.RECOGNIZE_TYPE_TEXT,
        Ocr.OUTPUT_FORMAT_PLAINTEXT);
    return text;
}
finally
{
    ocr.StopEngine();                         // Must call — no IDisposable fallback
}

// Asprise: four required steps before reading a single image
Ocr.SetUp();                              // Static global init
Ocr ocr = new Ocr();
try
{
    ocr.StartEngine("eng", Ocr.SPEED_FAST);   // Allocates native engine
    string text = ocr.Recognize(
        imagePath,
        Ocr.RECOGNIZE_TYPE_TEXT,
        Ocr.OUTPUT_FORMAT_PLAINTEXT);
    return text;
}
finally
{
    ocr.StopEngine();                         // Must call — no IDisposable fallback
}

' Asprise: four required steps before reading a single image
Ocr.SetUp()                              ' Static global init
Dim ocr As New Ocr()
Try
    ocr.StartEngine("eng", Ocr.SPEED_FAST)   ' Allocates native engine
    Dim text As String = ocr.Recognize( _
        imagePath, _
        Ocr.RECOGNIZE_TYPE_TEXT, _
        Ocr.OUTPUT_FORMAT_PLAINTEXT)
    Return text
Finally
    ocr.StopEngine()                         ' Must call — no IDisposable fallback
End Try

This pattern fails silently in three common situations: an unhandled exception before StopEngine() is called, a refactored code path that skips cleanup, and concurrent use on LITE/STANDARD where the license prohibits the threading that would even allow parallelism. The finally block mitigates the first issue, but the license restriction makes the pattern moot for server workloads anyway.

Detailed Feature Comparison

Asprise Approach

Asprise communicates between managed .NET code and its unmanaged OCR engine via P/Invoke marshaling. The asprise-vs-ironocr-examples.cs source shows the underlying mechanism:

// Asprise interop layer — bridging managed C# to native OCR engine
[DllImport("aocr.dll")]
private static extern IntPtr OCR(string imagePath, int type);

public string ExtractText(string imagePath)
{
    // P/Invoke call into unmanaged DLL
    IntPtr result = OCR(imagePath, 0);
    return Marshal.PtrToStringAnsi(result);  // Manual string marshal
}

// Asprise interop layer — bridging managed C# to native OCR engine
[DllImport("aocr.dll")]
private static extern IntPtr OCR(string imagePath, int type);

public string ExtractText(string imagePath)
{
    // P/Invoke call into unmanaged DLL
    IntPtr result = OCR(imagePath, 0);
    return Marshal.PtrToStringAnsi(result);  // Manual string marshal
}

Imports System.Runtime.InteropServices

' Asprise interop layer — bridging managed VB.NET to native OCR engine
Private Class OCRInterop
    <DllImport("aocr.dll")>
    Private Shared Function OCR(imagePath As String, type As Integer) As IntPtr
    End Function

    Public Function ExtractText(imagePath As String) As String
        ' P/Invoke call into unmanaged DLL
        Dim result As IntPtr = OCR(imagePath, 0)
        Return Marshal.PtrToStringAnsi(result)  ' Manual string marshal
    End Function
End Class

At the higher-level API, developers call into the same bridge through a class wrapper, but the deployment requirement does not change: every target machine needs the correct native binary, in the correct path, matching the correct process architecture. A 64-bit deployment that ships with aocr.dll instead of aocr_x64.dll throws a BadImageFormatException at runtime. A Linux Docker container without libaocr.so in LD_LIBRARY_PATH throws DllNotFoundException. Neither error is caught at build time.

IronOCR Approach

IronOCR installs as a single NuGet reference. The package handles all native dependencies internally through NuGet's platform-specific package selection mechanism:

// Installation — one command, all platforms
// dotnet add package IronOcr

// License setup
IronOcr.License.LicenseKey = "YOUR-LICENSE-KEY";

// Basic OCR — no native binary setup, no tessdata, no config files
var text = new IronTesseract().Read("document.jpg").Text;

// Installation — one command, all platforms
// dotnet add package IronOcr

// License setup
IronOcr.License.LicenseKey = "YOUR-LICENSE-KEY";

// Basic OCR — no native binary setup, no tessdata, no config files
var text = new IronTesseract().Read("document.jpg").Text;

Imports IronOcr

' Installation — one command, all platforms
' dotnet add package IronOcr

' License setup
IronOcr.License.LicenseKey = "YOUR-LICENSE-KEY"

' Basic OCR — no native binary setup, no tessdata, no config files
Dim text As String = New IronTesseract().Read("document.jpg").Text

The using statement on OcrInput replaces manual engine lifecycle calls. There is no StartEngine() to call and no StopEngine() to forget. The garbage collector and IDisposable pattern handle cleanup correctly in exception scenarios without requiring try/finally scaffolding around every OCR call.

For detailed setup guidance, see the IronTesseract setup guide.

Asprise Approach

LITE and STANDARD licenses explicitly prohibit multi-threaded and multi-process execution. Any code that calls Asprise from more than one thread simultaneously violates the license agreement on those tiers. ASP.NET Core processes requests on a thread pool by default. That makes every standard Web API controller that calls Asprise a license violation on LITE/STANDARD:

// ASPRISE LITE/STANDARD — violates license in ASP.NET Core context
// Web server thread pool = multiple concurrent threads = prohibited
[ApiController]
public class OcrController : ControllerBase
{
    [HttpPost("extract")]
    public IActionResult ExtractText(IFormFile file)
    {
        // Two concurrent requests = two threads = license violation
        var ocr = new Ocr();
        ocr.StartEngine("eng", Ocr.SPEED_FAST);
        var text = ocr.Recognize(tempPath, Ocr.RECOGNIZE_TYPE_TEXT, Ocr.OUTPUT_FORMAT_PLAINTEXT);
        ocr.StopEngine();
        return Ok(text);
    }
}

// ASPRISE LITE/STANDARD — violates license in ASP.NET Core context
// Web server thread pool = multiple concurrent threads = prohibited
[ApiController]
public class OcrController : ControllerBase
{
    [HttpPost("extract")]
    public IActionResult ExtractText(IFormFile file)
    {
        // Two concurrent requests = two threads = license violation
        var ocr = new Ocr();
        ocr.StartEngine("eng", Ocr.SPEED_FAST);
        var text = ocr.Recognize(tempPath, Ocr.RECOGNIZE_TYPE_TEXT, Ocr.OUTPUT_FORMAT_PLAINTEXT);
        ocr.StopEngine();
        return Ok(text);
    }
}

Imports Microsoft.AspNetCore.Mvc

' ASPRISE LITE/STANDARD — violates license in ASP.NET Core context
' Web server thread pool = multiple concurrent threads = prohibited
<ApiController>
Public Class OcrController
    Inherits ControllerBase

    <HttpPost("extract")>
    Public Function ExtractText(file As IFormFile) As IActionResult
        ' Two concurrent requests = two threads = license violation
        Dim ocr As New Ocr()
        ocr.StartEngine("eng", Ocr.SPEED_FAST)
        Dim text As String = ocr.Recognize(tempPath, Ocr.RECOGNIZE_TYPE_TEXT, Ocr.OUTPUT_FORMAT_PLAINTEXT)
        ocr.StopEngine()
        Return Ok(text)
    End Function
End Class

Batch processing on LITE/STANDARD is forced sequential regardless of available CPU cores:

// ASPRISE LITE/STANDARD — sequential only (100 docs at 2 sec each = 3+ minutes)
Ocr.SetUp();
Ocr ocr = new Ocr();
try
{
    ocr.StartEngine("eng", Ocr.SPEED_FAST);
    foreach (var path in imagePaths)  // Cannot use Parallel.ForEach — license violation
    {
        string text = ocr.Recognize(path, Ocr.RECOGNIZE_TYPE_TEXT, Ocr.OUTPUT_FORMAT_PLAINTEXT);
        results.Add(text);
    }
}
finally
{
    ocr.StopEngine();
}

// ASPRISE LITE/STANDARD — sequential only (100 docs at 2 sec each = 3+ minutes)
Ocr.SetUp();
Ocr ocr = new Ocr();
try
{
    ocr.StartEngine("eng", Ocr.SPEED_FAST);
    foreach (var path in imagePaths)  // Cannot use Parallel.ForEach — license violation
    {
        string text = ocr.Recognize(path, Ocr.RECOGNIZE_TYPE_TEXT, Ocr.OUTPUT_FORMAT_PLAINTEXT);
        results.Add(text);
    }
}
finally
{
    ocr.StopEngine();
}

' ASPRISE LITE/STANDARD — sequential only (100 docs at 2 sec each = 3+ minutes)
Ocr.SetUp()
Dim ocr As New Ocr()
Try
    ocr.StartEngine("eng", Ocr.SPEED_FAST)
    For Each path In imagePaths ' Cannot use Parallel.ForEach — license violation
        Dim text As String = ocr.Recognize(path, Ocr.RECOGNIZE_TYPE_TEXT, Ocr.OUTPUT_FORMAT_PLAINTEXT)
        results.Add(text)
    Next
Finally
    ocr.StopEngine()
End Try

ENTERPRISE removes the thread restriction, but ENTERPRISE requires contacting Asprise sales with no published price.

IronOCR Approach

IronTesseract is thread-safe and carries no threading restrictions on any license tier. Parallel batch processing on a $999 Lite license is fully supported:

// IronOCR — parallel batch on any license tier
// 100 docs at 2 sec each, 8 cores = ~25 seconds vs 200 seconds sequential
var results = imagePaths
    .AsParallel()
    .Select(path => new IronTesseract().Read(path).Text)
    .ToList();

// IronOCR — parallel batch on any license tier
// 100 docs at 2 sec each, 8 cores = ~25 seconds vs 200 seconds sequential
var results = imagePaths
    .AsParallel()
    .Select(path => new IronTesseract().Read(path).Text)
    .ToList();

Imports System.Linq

' IronOCR — parallel batch on any license tier
' 100 docs at 2 sec each, 8 cores = ~25 seconds vs 200 seconds sequential
Dim results = imagePaths _
    .AsParallel() _
    .Select(Function(path) (New IronTesseract()).Read(path).Text) _
    .ToList()

ASP.NET Core controllers work without any special configuration:

// IronOCR — concurrent requests on any license tier
[ApiController]
public class OcrController : ControllerBase
{
    [HttpPost("extract")]
    public IActionResult ExtractText(IFormFile file)
    {
        // Thread-safe on Lite, Plus, Professional, Unlimited — all tiers
        var text = new IronTesseract().Read(tempPath).Text;
        return Ok(text);
    }
}

// IronOCR — concurrent requests on any license tier
[ApiController]
public class OcrController : ControllerBase
{
    [HttpPost("extract")]
    public IActionResult ExtractText(IFormFile file)
    {
        // Thread-safe on Lite, Plus, Professional, Unlimited — all tiers
        var text = new IronTesseract().Read(tempPath).Text;
        return Ok(text);
    }
}

Imports Microsoft.AspNetCore.Mvc

<ApiController>
Public Class OcrController
    Inherits ControllerBase

    <HttpPost("extract")>
    Public Function ExtractText(file As IFormFile) As IActionResult
        ' Thread-safe on Lite, Plus, Professional, Unlimited — all tiers
        Dim text = New IronTesseract().Read(tempPath).Text
        Return Ok(text)
    End Function
End Class

See the multithreading example for parallel throughput patterns in batch workloads.

Asprise Approach

Preprocessing requires an external image library. A developer working with Asprise adds a dependency like ImageMagick, SkiaSharp, or System.Drawing, runs the preprocessing operations manually, saves the processed image to a temporary file, and then passes that file to Asprise:

// Asprise preprocessing: external dependency required
// 1. Load with external library
// 2. Apply corrections (deskew, denoise, contrast) with external library
// 3. Save to temp file
// 4. Pass temp file to Asprise

var ocr = new Ocr();
ocr.StartEngine("eng", Ocr.SPEED_FAST);
// preprocessedImagePath comes from your external preprocessing pipeline
string text = ocr.Recognize(preprocessedImagePath, Ocr.RECOGNIZE_TYPE_TEXT, Ocr.OUTPUT_FORMAT_PLAINTEXT);
ocr.StopEngine();

// Asprise preprocessing: external dependency required
// 1. Load with external library
// 2. Apply corrections (deskew, denoise, contrast) with external library
// 3. Save to temp file
// 4. Pass temp file to Asprise

var ocr = new Ocr();
ocr.StartEngine("eng", Ocr.SPEED_FAST);
// preprocessedImagePath comes from your external preprocessing pipeline
string text = ocr.Recognize(preprocessedImagePath, Ocr.RECOGNIZE_TYPE_TEXT, Ocr.OUTPUT_FORMAT_PLAINTEXT);
ocr.StopEngine();

' Asprise preprocessing: external dependency required
' 1. Load with external library
' 2. Apply corrections (deskew, denoise, contrast) with external library
' 3. Save to temp file
' 4. Pass temp file to Asprise

Dim ocr As New Ocr()
ocr.StartEngine("eng", Ocr.SPEED_FAST)
' preprocessedImagePath comes from your external preprocessing pipeline
Dim text As String = ocr.Recognize(preprocessedImagePath, Ocr.RECOGNIZE_TYPE_TEXT, Ocr.OUTPUT_FORMAT_PLAINTEXT)
ocr.StopEngine()

This adds a build dependency, adds runtime overhead from the extra library, and requires the developer to understand image processing well enough to implement effective corrections.

IronOCR Approach

Preprocessing is built into OcrInput. The same pipeline that took 50-100 lines with an external library and careful parameter tuning becomes five method calls:

// IronOCR — preprocessing built in, no external dependencies
using var input = new OcrInput();
input.LoadImage("low-quality-scan.jpg");
input.Deskew();               // Correct page rotation
input.DeNoise();              // Remove scanner artifacts
input.Contrast();             // Improve text/background separation
input.Binarize();             // Convert to black/white for cleaner engine input
input.EnhanceResolution(300); // Scale to optimal DPI

var result = new IronTesseract().Read(input);
Console.WriteLine($"Confidence: {result.Confidence}%");

// IronOCR — preprocessing built in, no external dependencies
using var input = new OcrInput();
input.LoadImage("low-quality-scan.jpg");
input.Deskew();               // Correct page rotation
input.DeNoise();              // Remove scanner artifacts
input.Contrast();             // Improve text/background separation
input.Binarize();             // Convert to black/white for cleaner engine input
input.EnhanceResolution(300); // Scale to optimal DPI

var result = new IronTesseract().Read(input);
Console.WriteLine($"Confidence: {result.Confidence}%");

Imports IronOcr

' IronOCR — preprocessing built in, no external dependencies
Using input As New OcrInput()
    input.LoadImage("low-quality-scan.jpg")
    input.Deskew()               ' Correct page rotation
    input.DeNoise()              ' Remove scanner artifacts
    input.Contrast()             ' Improve text/background separation
    input.Binarize()             ' Convert to black/white for cleaner engine input
    input.EnhanceResolution(300) ' Scale to optimal DPI

    Dim result = New IronTesseract().Read(input)
    Console.WriteLine($"Confidence: {result.Confidence}%")
End Using

For scans where rotation is unpredictable, automatic deskew handles arbitrary angles without requiring the developer to detect or specify rotation values. The image quality correction guide covers the full filter set and when to apply each.

Asprise Approach

// Asprise PDF workaround — external library required to render PDF pages
var images = ExternalPdfLibrary.RenderPages("invoice.pdf");

Ocr.SetUp();
Ocr ocr = new Ocr();
try
{
    ocr.StartEngine("eng", Ocr.SPEED_FAST);
    var allText = new System.Text.StringBuilder();
    foreach (var imagePath in images)
    {
        string pageText = ocr.Recognize(imagePath, Ocr.RECOGNIZE_TYPE_TEXT, Ocr.OUTPUT_FORMAT_PLAINTEXT);
        allText.Append(pageText);
    }
    return allText.ToString();
}
finally
{
    ocr.StopEngine();
}
// Also: clean up temp image files

// Asprise PDF workaround — external library required to render PDF pages
var images = ExternalPdfLibrary.RenderPages("invoice.pdf");

Ocr.SetUp();
Ocr ocr = new Ocr();
try
{
    ocr.StartEngine("eng", Ocr.SPEED_FAST);
    var allText = new System.Text.StringBuilder();
    foreach (var imagePath in images)
    {
        string pageText = ocr.Recognize(imagePath, Ocr.RECOGNIZE_TYPE_TEXT, Ocr.OUTPUT_FORMAT_PLAINTEXT);
        allText.Append(pageText);
    }
    return allText.ToString();
}
finally
{
    ocr.StopEngine();
}
// Also: clean up temp image files

Imports System.Text

' Asprise PDF workaround — external library required to render PDF pages
Dim images = ExternalPdfLibrary.RenderPages("invoice.pdf")

Ocr.SetUp()
Dim ocr As New Ocr()
Try
    ocr.StartEngine("eng", Ocr.SPEED_FAST)
    Dim allText As New StringBuilder()
    For Each imagePath In images
        Dim pageText As String = ocr.Recognize(imagePath, Ocr.RECOGNIZE_TYPE_TEXT, Ocr.OUTPUT_FORMAT_PLAINTEXT)
        allText.Append(pageText)
    Next
    Return allText.ToString()
Finally
    ocr.StopEngine()
End Try
' Also: clean up temp image files

This approach adds a PDF rendering dependency (typically iText, PDFSharp, or a commercial renderer), requires managing temporary files, and loses PDF metadata and structure that a native PDF reader would preserve. Password-protected PDFs require a third component.

IronOCR Approach

IronOCR reads PDFs directly. The OcrInput.LoadPdf() method handles rendering internally, including multi-page documents and password-protected files:

// IronOCR — native PDF input, no external rendering library
using var input = new OcrInput();
input.LoadPdf("invoice.pdf");
var result = new IronTesseract().Read(input);
Console.WriteLine(result.Text);

// Password-protected PDFs — same API, one parameter
using var secureInput = new OcrInput();
secureInput.LoadPdf("confidential.pdf", Password: "secret");
var secureResult = new IronTesseract().Read(secureInput);

// Generate searchable PDF from a scanned document
var scanResult = new IronTesseract().Read("scanned-contract.pdf");
scanResult.SaveAsSearchablePdf("searchable-contract.pdf");

// IronOCR — native PDF input, no external rendering library
using var input = new OcrInput();
input.LoadPdf("invoice.pdf");
var result = new IronTesseract().Read(input);
Console.WriteLine(result.Text);

// Password-protected PDFs — same API, one parameter
using var secureInput = new OcrInput();
secureInput.LoadPdf("confidential.pdf", Password: "secret");
var secureResult = new IronTesseract().Read(secureInput);

// Generate searchable PDF from a scanned document
var scanResult = new IronTesseract().Read("scanned-contract.pdf");
scanResult.SaveAsSearchablePdf("searchable-contract.pdf");

Imports IronOcr

' IronOCR — native PDF input, no external rendering library
Using input As New OcrInput()
    input.LoadPdf("invoice.pdf")
    Dim result = New IronTesseract().Read(input)
    Console.WriteLine(result.Text)
End Using

' Password-protected PDFs — same API, one parameter
Using secureInput As New OcrInput()
    secureInput.LoadPdf("confidential.pdf", Password:="secret")
    Dim secureResult = New IronTesseract().Read(secureInput)
End Using

' Generate searchable PDF from a scanned document
Dim scanResult = New IronTesseract().Read("scanned-contract.pdf")
scanResult.SaveAsSearchablePdf("searchable-contract.pdf")

The PDF OCR guide covers page range selection, handling mixed PDF types, and searchable PDF output for archival workflows.

The Production Readiness Blocker

The thread restriction surfaces as a hard blocker for most production .NET workloads. A team builds an OCR feature using Asprise LITE during development — single-threaded testing passes, everything works. Then the feature ships to a staging environment behind an ASP.NET Core API, concurrent test requests arrive, and the application either throws errors or operates in a technically non-compliant state. The resolution is either upgrading to ENTERPRISE (with its associated cost and sales engagement) or replacing the library. Teams that reach this point during staging rather than in production are fortunate; those who discover it post-launch face a more urgent migration. IronOCR removes this entire class of problem — the $999 Lite license supports concurrent request processing, batch parallelism, Windows Services, and cloud functions without restriction.

The Deployment Complexity Tax

Teams operating in modern deployment environments — Docker containers, Linux VMs, Kubernetes pods — pay an ongoing maintenance cost with Asprise that does not exist with pure NuGet packages. Every container image build must include the correct native binary for the target architecture. Every CI/CD pipeline that targets multiple platforms must handle platform-specific file inclusion. A missing native library on a production Linux container is a runtime discovery, not a build-time error. Teams that have spent hours debugging a 32-bit DLL loaded into a 64-bit process understand the cost concretely. IronOCR's NuGet-only deployment eliminates this entire category of deployment failure.

The Java Documentation Problem

Teams building .NET applications need C# examples, not Java examples. Asprise's primary documentation, API references, and community resources are oriented toward Java developers. A .NET developer reading Asprise documentation translates Java syntax, Java package conventions, and Java-specific patterns into C# equivalents — and sometimes the translation is not direct because the .NET wrapper does not expose every Java-side feature. IronOCR's documentation, tutorials, and code samples are written for C# developers. The reading text from images tutorial and the full tutorials hub provide working C# examples for every feature without translation overhead.

The PDF Pipeline Gap

Teams processing scanned documents, invoices, contracts, or forms in PDF format cannot use Asprise without adding a PDF rendering library to their dependency graph. That rendering library introduces its own licensing considerations, maintenance burden, and potential incompatibilities. For teams that need OCR + PDF in a single stack, IronOCR handles both natively. The ability to create searchable PDFs from scanned input — a common requirement for archival and compliance workflows — has no equivalent in Asprise at any license tier.

The Language Coverage Gap

Asprise supports 20+ OCR languages. IronOCR supports 125+, each installable as an independent NuGet package. Teams processing documents in Arabic, Hindi, Thai, or any of dozens of other languages supported by IronOCR but not by Asprise have no path to multilingual support through Asprise. The multiple languages guide covers installing and combining language packs, including simultaneous recognition across multiple scripts.

Replacing Engine Lifecycle Code

The largest mechanical change in migration is removing the Asprise engine lifecycle pattern and replacing it with direct IronTesseract instantiation. Every occurrence of the SetUp() / StartEngine() / Recognize() / StopEngine() sequence becomes a single Read() call:

// Before: Asprise lifecycle (15 lines, manual cleanup)
Ocr.SetUp();
Ocr ocr = new Ocr();
try
{
    ocr.StartEngine("eng", Ocr.SPEED_FAST);
    string text = ocr.Recognize(
        imagePath,
        Ocr.RECOGNIZE_TYPE_TEXT,
        Ocr.OUTPUT_FORMAT_PLAINTEXT);
    return text;
}
finally
{
    ocr.StopEngine();
}

// After: IronOCR (1 line)
return new IronTesseract().Read(imagePath).Text;

// Before: Asprise lifecycle (15 lines, manual cleanup)
Ocr.SetUp();
Ocr ocr = new Ocr();
try
{
    ocr.StartEngine("eng", Ocr.SPEED_FAST);
    string text = ocr.Recognize(
        imagePath,
        Ocr.RECOGNIZE_TYPE_TEXT,
        Ocr.OUTPUT_FORMAT_PLAINTEXT);
    return text;
}
finally
{
    ocr.StopEngine();
}

// After: IronOCR (1 line)
return new IronTesseract().Read(imagePath).Text;

' Before: Asprise lifecycle (15 lines, manual cleanup)
Ocr.SetUp()
Dim ocr As New Ocr()
Try
    ocr.StartEngine("eng", Ocr.SPEED_FAST)
    Dim text As String = ocr.Recognize( _
        imagePath, _
        Ocr.RECOGNIZE_TYPE_TEXT, _
        Ocr.OUTPUT_FORMAT_PLAINTEXT)
    Return text
Finally
    ocr.StopEngine()
End Try

' After: IronOCR (1 line)
Return New IronTesseract().Read(imagePath).Text

For performance-sensitive code that processes many documents sequentially, reuse the IronTesseract instance rather than creating a new one per call — engine initialization carries overhead, and a single instance used sequentially is more efficient than creating and disposing one per document.

Replacing String-Based Language Codes

Asprise uses string parameters for language selection ("eng", "fra", "eng+fra"). IronOCR uses a strongly-typed OcrLanguage enum with a secondary language API. The multiple languages guide covers available language identifiers and the NuGet packages required for each:

// Before: Asprise string-based language
ocr.StartEngine("eng+fra", Ocr.SPEED_FAST);

// After: IronOCR strongly-typed language enum
var ocr = new IronTesseract();
ocr.Language = OcrLanguage.English;
ocr.AddSecondaryLanguage(OcrLanguage.French);
var result = ocr.Read(imagePath);

// Before: Asprise string-based language
ocr.StartEngine("eng+fra", Ocr.SPEED_FAST);

// After: IronOCR strongly-typed language enum
var ocr = new IronTesseract();
ocr.Language = OcrLanguage.English;
ocr.AddSecondaryLanguage(OcrLanguage.French);
var result = ocr.Read(imagePath);

Imports IronOcr

' Before: Asprise string-based language
ocr.StartEngine("eng+fra", Ocr.SPEED_FAST)

' After: IronOCR strongly-typed language enum
Dim ocr As New IronTesseract()
ocr.Language = OcrLanguage.English
ocr.AddSecondaryLanguage(OcrLanguage.French)
Dim result = ocr.Read(imagePath)

Replacing Error-String Checks

Asprise returns null or an error-prefixed string when recognition fails. IronOCR throws .NET exceptions. Replace null and string-prefix checks with standard try/catch blocks. This aligns OCR error handling with the rest of the .NET exception handling model and gives access to full stack traces and exception type hierarchies rather than parsed error strings.

Enabling Parallel Processing

After migration, existing foreach batch loops can be converted to Parallel.ForEach or PLINQ without any licensing concern. For document batches that previously processed sequentially on LITE/STANDARD, parallelism is a direct throughput multiplier. The async OCR guide covers async patterns for web application contexts where non-blocking OCR is preferable to thread pool saturation.