Skip to content

Tesseract wrapper for Windows, Android and iOS for .NET MAUI

License

Notifications You must be signed in to change notification settings

henrivain/TesseractOcrMaui

Repository files navigation

Deploy to NuGet Deploy iOS package to NuGet

TesseractOcrMaui

Tesseract wrapper for Windows, iOS and Android for .NET MAUI.

What is this?

.NET MAUI wrapper for optical character recognizion library Tesseract. I didn't find any good up to date C# wrapper to Tesseract that would function with Maui on Android devices. This library wrapps native Tesseract C/C++ libraries to usable C# interfaces.

Supported platforms

Currently supports Windows, iOS and Android. Library is meant to be used with .NET MAUI project. You can see supported cpu architechtures down below.

platform Architechture
Windows x86_64
Android Arm64-v8a
Android Arm-v7a
Android x86_64
Android x86
iOS emulator x86_64
iOS emulator Arm64
iOS physical device Arm64

Supported runtimes

net7.0 or newer
net7.0-windows10.0.19041 or newer
net7.0-android or newer
net7.0-ios or newer

Only png and jpeg libraries are compiled into tesseract native libraries, so only these image types are supported. Additional image libraries are added if needed later.

Getting started

1. Add nuget package to your project

Package is available in nuget.org. You can use it in your project by adding it in your :

  1. Visual Studio Nuget Package Manager Search TesseractOcrMaui and add it to your Maui project

  2. Using Dotnet CLI run command

dotnet add package TesseractOcrMaui
  1. By package reference
<PackageReference Include="TesseractOcrMaui" Version="1.2.0" />

2. Add package to dependency injection (see TesseractOcrMauiTestApp)

Add AddTesseractOcr call in you injections. This loads all libraries that are needed to run library and adds needed injectable interfaces. files.AddFile("eng.traineddata"); adds all your traineddata files to be loaded, when tesseract is used. For example I add eng.traineddata, so I must add traineddata file with same name to my project's TesseractOcrMauiTestApp\Resources\Raw folder.

MauiProgram.cs

using Microsoft.Extensions.Logging;
using TesseractOcrMaui;  // include library namespace

namespace TesseractOcrMauiTestApp;

public static class MauiProgram
{
    public static MauiApp CreateMauiApp()
    {
        var builder = MauiApp.CreateBuilder();
        builder
            .UseMauiApp<App>()
            .ConfigureFonts(fonts =>
            {
                fonts.AddFont("OpenSans-Regular.ttf", "OpenSansRegular");
                fonts.AddFont("OpenSans-Semibold.ttf", "OpenSansSemibold");
            });
        // Inject logging, (optional, but might give useful information)
        builder.Services.AddLogging();

        // Inject library functionality
        builder.Services.AddTesseractOcr(
            files =>
            {
                // must have matching files in Resources/Raw folder
                files.AddFile("eng.traineddata");
            });

        // Inject main page, so services are injected to its constructor
        builder.Services.AddSingleton<MainPage>();

        return builder.Build();
    }
}

3. Inject ITesseract into your page

Now you can constructor inject ITesseract interface to you page. I have two labels ("confidenceLabel" and "resultLabel") in my main page. I added button with clicked event handler. You can see my Button_Clicked functionality down below.

Mainpage.xaml.cs

using TesseractOcrMaui; // Include library namespace
using TesseractOcrMaui.Extensions;  // Help for handling result object from recognizion

namespace TesseractOcrMauiTestApp;

public partial class MainPage : ContentPage
{
    // inject ITesseract
    public MainPage(ITesseract tesseract)
    {
        InitializeComponent();
        Tesseract = tesseract;
    }

    ITesseract Tesseract { get; }

    private async void Button_Clicked(object sender, EventArgs e)
    {
        // Make user pick image path
        var pickResult = await GetUserSelectedImagePath();

        // null if user cancelled the operation
        if (pickResult is null)
        {
            return;
        }

        // Recognize image
        var result = await Tesseract.RecognizeTextAsync(pickResult.FullPath);

        // Show output
        confidenceLabel.Text = $"Confidence: {result.Confidence}";
        if (result.NotSuccess())
        {
            resultLabel.Text = $"Recognizion failed: {result.Status}";
            return;
        }
        resultLabel.Text = result.RecognisedText;
    }

    private static async Task<string?> GetUserSelectedImagePath()
    {
        /* This method lets user to select image file by opening
           file selection dialog. */
#if IOS
        /* Note that this method uses conditional
           compilation for iOS because MediaPicker is better
           option to use for image picking on iOS. */
        var pickResult = await MediaPicker.PickPhotoAsync(new MediaPickerOptions()
        {
            Title = "Pick jpeg or png image"
        });
#else
        var pickResult = await FilePicker.PickAsync(new PickOptions()
        {
            PickerTitle = "Pick jpeg or png image",
            // Currently usable image types are png and jpeg
            FileTypes = new FilePickerFileType(new Dictionary<DevicePlatform, IEnumerable<string>>()
            {
                [DevicePlatform.Android] = new List<string>() { "image/png", "image/jpeg" },
                [DevicePlatform.WinUI] = new List<string>() { ".png", ".jpg", ".jpeg" },
            })
        });
#endif
        return pickResult?.FullPath;
    }
}

ITesseract API

You can find following methods in the main high level API.

public interface ITesseract
{
    RecognizionResult RecognizeText(string imagePath);
    RecognizionResult RecognizeText(byte[] imageBytes);
    RecognizionResult RecognizeText(Pix image);

    Task<RecognizionResult> RecognizeTextAsync(string imagePath);
    Task<RecognizionResult> RecognizeTextAsync(byte[] imageBytes);
    Task<RecognizionResult> RecognizeTextAsync(Pix pix);

    // Loads traineddata files for use from app packages to appdata folder
    Task<DataLoadResult> LoadTraineddataAsync();

    // Gets tessdata folder path from TessDataProvider (from configuration)
    string TessDataFolder { get; }

    // Sets tesseract engine to use for example lstm model or tesseract only
    EngineMode EngineMode { get; set; }

    // Access used TessEngine for configuration (E.g. Whitelist chafacters)
    Action<ITessEngineConfigurable>? EngineConfiguration { set; }
}

Licence

Copyright 2023 Henri Vainio

Licensed under the Apache License, Version 2.0 (the "License");
you may not use this file except in compliance with the License.
You may obtain a copy of the License at

http://www.apache.org/licenses/LICENSE-2.0

Unless required by applicable law or agreed to in writing, software
distributed under the License is distributed on an "AS IS" BASIS,
WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
See the License for the specific language governing permissions and
limitations under the License.

NOTE: Tesseract depends on other packages that may be licensed under different open source licenses.

This project does not depend on any third-party C# packages, but it needs traineddata files to function. Parts of the code are also reused from Charlesw Windows Tesseract wrapper.

Documentation

Maccatalyst support

Maccatalyst is not currently supported, but the support is probably added in future. If you are interested to get the Maccatalyst support sooner, create a github issue or send me an email. I cannot develop for Maccatalyst myself, so it is greatly dependant on outside contributors.

Thanks to contributors

Ivan Mendoza @sivanmg - iOS support

Support

If you have any questions about anything related to this project, open a issue on Github or send me an email. If something is not working for you or you have a feature request, just, let me know!

Henri Vainio
matikkaeditorinkaantaja(at)gmail.com