FaceRecognization: A C# repository from TechnologyK

Face Recognization

This project is a C# wrapper of arcsoft's library.

Stepon.FaceRecognizationCore is for .net core 2.0. It uses CoreCompat/System.Drawing, you should install libgdiplus.

How to use

Nuget Package

For .net framework 4.5.1 (V1.1.0):

Install-Package Stepon.FaceRecognization

For .net standard 2.0 (V1.1.0):

Install-Package Stepon.FaceRecognizationCore

Get app id and sdk key from arcsoft's website (it's free). Remember to update the libraries if arcsoft has newer version. The appid and key is associated with sdk.

There are three main classes to use for recognization:

FaceDetection for image to detect face location
FaceTracking for video to detect face location
FaceRecognize for face's feature extracting and matching
FaceAge for age estimation
FaceGender for gender estimation

FaceProcessor is a high level wrapper for above three classes. You can just use this class to complete the face verify task.

Example

Face detecting:

using (var detection = LocatorFactory.GetDetectionLocator("appId", "sdkKey"))
{
    var image = Image.FromFile("test.jpg");
    var bitmap = new Bitmap(image);

    var result = detection.Detect(bitmap, out var locateResult);
    using (locateResult)
    {
        if (result == ErrorCode.Ok && locateResult.FaceCount > 0)
        {
            using (var g = Graphics.FromImage(bitmap))
            {
                var face = locateResult.Faces[0].ToRectangle();
                g.DrawRectangle(new Pen(Color.Chartreuse), face.X, face.Y, face.Width, face.Height);
            }

            bitmap.Save("output.jpg", ImageFormat.Jpeg);
        }
    }
}

Face traking (recommended for video face traking, this just an example for image, check FaceDemo for complete example):

using (var detection = LocatorFactory.GetTrackingLocator("appId", "sdkKey"))
{
    var image = Image.FromFile("test.jpg");
    var bitmap = new Bitmap(image);

    var result = detection.Detect(bitmap, out var locateResult);
    using (locateResult)
    {
        if (result == ErrorCode.Ok && locateResult.FaceCount > 0)
        {
            using (var g = Graphics.FromImage(bitmap))
            {
                var face = locateResult.Faces[0].ToRectangle();
                g.DrawRectangle(new Pen(Color.Chartreuse), face.X, face.Y, face.Width, face.Height);
            }

            bitmap.Save("output.jpg", ImageFormat.Jpeg);
        }
    }
}

Face matching:

using (var proccesor = new FaceProcessor("appid",
                "locatorKey", "recognizeKey", true))
{
    var image1 = Image.FromFile("test2.jpg");
    var image2 = Image.FromFile("test.jpg");

    var result1 = proccesor.LocateExtract(new Bitmap(image1));
    var result2 = proccesor.LocateExtract(new Bitmap(image2));
    
    //you can persist result1[0].FeatureData to further use

    if ((result1 != null) & (result2 != null))
        Console.WriteLine(proccesor.Match(result1[0].FeatureData, result2[0].FeatureData, true));
}

Face age and gender estimate:

var age = new FaceAge("appid", "key");
var gender = new FaceGender("appid", "key");
using (var detection = LocatorFactory.GetDetectionLocator("appid", "key", age, gender))
{
    var image1 = Image.FromFile("test.jpg");
    var bitmap = new Bitmap(image1);
    var result = detection.Detect(bitmap, out var location,
        LocateOperation.IncludeAge | LocateOperation.IncludeGender);//default is None, no age and gender estimate

    using (var g = Graphics.FromImage(bitmap))
    {
        for (var i = 0; i < location.FaceCount; i++)
        {
            var face = location.Faces[i].ToRectangle();
            g.DrawRectangle(new Pen(Color.Chartreuse), face.X, face.Y, face.Width, face.Height);
            g.DrawString($"age:{location.Ages[i]},gender:{location.Genders[i]}", new Font("Yanhei", 12), new SolidBrush(Color.Aqua), face.X, face.Y);
        } 
    }

    bitmap.Save("Ok.jpg", ImageFormat.Jpeg);

    location.Dispose();
}
age.Dispose();
gender.Dispose();

It also can be used standalone:

using (var detection = LocatorFactory.GetDetectionLocator("appid", "key")) // should pay attension of static or preview mode, you should use detection for static estimation and tracking for prview estimation
{
    var image1 = Image.FromFile("test2.jpg");
    using (var estimate = new FaceAge("appid", "key"))
    {
        var result1 = estimate.StaticEstimation(detection, new Bitmap(image1));
        foreach (var result1Age in result1.Ages)
        {
            Console.WriteLine(result1Age);
        }
    }

    using (var estimate = new FaceGender("appid", "key"))
    {
        var result1 = estimate.StaticEstimation(detection, new Bitmap(image1));
        foreach (var result1Gender in result1.Genders)
        {
            Console.WriteLine(result1Gender);
        }
    }
}

FaceDemo

This is a complete example to show the using of this library. It use ffmepg(NReco.VideoConverter) to capture IP camera via RTSP. You can also use Emgu.CV(or other library) to capture web camera.

TechnologyK/FaceRecognization