Using Custom OCR Language Packs with IronOCR

How to create custom language packs for use in IronOCR?

Creating a custom language pack requires training a new Tesseract 4 LSTM language file/dictionary from a font.

There are many tutorials available online explaining the steps required to do this. The process is not simple, but it is thankfully quite well-documented.

As a good place to start, we suggest this YouTube tutorial from Gabriel Garcia (no affiliation) and their linked GitHub repository.

Once complete, the output will be a .traineddata file.

The .traineddata file can then be referenced in IronOCR as follows:

Documentation: IronOCR Custom Languages

using IronOcr;

class Program
{
    static void Main()
    {
        // Initialize the IronTesseract OCR engine
        var Ocr = new IronTesseract();

        // Load your custom Tesseract language file (trained .traineddata file)
        Ocr.UseCustomTesseractLanguageFile("mydir/custom.traineddata");  //<--- your new font

        // Multiple fonts can be used by calling the method multiple times with different files

        // Load an image into the OCR Input for processing
        using (var Input = new OcrInput(@"images\image.png"))
        {
            // Perform OCR on the input image
            var Result = Ocr.Read(Input);

            // Output the recognized text to the console
            Console.WriteLine(Result.Text);
        }
    }
}

using IronOcr;

class Program
{
    static void Main()
    {
        // Initialize the IronTesseract OCR engine
        var Ocr = new IronTesseract();

        // Load your custom Tesseract language file (trained .traineddata file)
        Ocr.UseCustomTesseractLanguageFile("mydir/custom.traineddata");  //<--- your new font

        // Multiple fonts can be used by calling the method multiple times with different files

        // Load an image into the OCR Input for processing
        using (var Input = new OcrInput(@"images\image.png"))
        {
            // Perform OCR on the input image
            var Result = Ocr.Read(Input);

            // Output the recognized text to the console
            Console.WriteLine(Result.Text);
        }
    }
}