Vitruvius is a set of easy-to-use Kinect utilities that will speed-up the development of your projects. Supports gesture detection, skeleton drawing, frame capturing, voice recognition and much more.

Great news for early-access developers: Vitruvius now supports Kinect for Windows version 2 sensor!

• Joint scaling, proper for on-screen display
• User height
• Distance between joints
• One-line skeleton tracking

• Easily display color, depth and infrared frames
• Save Kinect frames as bitmap images
• One-line skeleton drawing (WPF-only)

• WaveLeft
• WaveRight
• SwipeLeft
• SwipeRight
• JoinedHands
• SwipeUp
• SwipeDown
• ZoomIn
• ZoomOut
• Menu

• Recognize voice commands
• Speech synthesis

• Posture support (jumping, dancing, etc)

• Kinect for Windows or Kinect for XBOX sensor
• Kinect for Windows SDK v1.8

If you are developing using the developer preview Kinect SDK v2, you need the appropriate hardware and software provided by Microsoft.

• Download project's source code and build the solution that matches the version of your sensor. Version 2 refers to the private Developer Preview Kinect for Windows sensor.

1. Displaying Kinect color frames:

    void Sensor_ColorFrameReady(object sender, ColorImageFrameReadyEventArgs e)
    {
        using (var frame = e.OpenColorImageFrame())
        {
            if (frame != null)
            {
                // Display on screen
                image.Source = frame.ToBitmap();
                
                // Capture JPEG file
                frame.Capture("C:\\ColorFrame.jpg");
            }
        }
    }
   

2. Displaying Kinect depth frames:

    void Sensor_DepthFrameReady(object sender, DepthImageFrameReadyEventArgs e)
    {
        using (var frame = e.OpenDepthImageFrame())
        {
            if (frame != null)
            {
                // Display on screen
                image.Source = frame.ToBitmap();
                
                // Capture JPEG file
                frame.Capture("C:\\DepthFrame.jpg");
            }
        }
    }
   

3. Drawing a skeleton and getting its height:

    void Sensor_SkeletonFrameReady(object sender, SkeletonFrameReadyEventArgs e)
    {
        using (var frame = e.OpenSkeletonFrame())
        {
            if (frame != null)
            {
                canvas.ClearSkeletons();
                
                var skeletons = frame.Skeletons().Where(s => s.TrackingState == SkeletonTrackingState.Tracked);
                
                foreach (var skeleton in skeletons)
                {
                    if (skeleton != null)
                    {
                        // Draw the skeleton.
                        canvas.DrawSkeleton(skeleton);
                            
                        // Get the skeleton height.
                        double height = skeleton.Height();
                    }
                }
            }
        }
    }
   

4. Detecting gestures:

    GestureController gestureController = new GestureController(GestureType.All);
    gestureController.GestureRecognized += GestureController_GestureRecognized;
    
    // ...
    
    void Sensor_SkeletonFrameReady(object sender, SkeletonFrameReadyEventArgs e)
    {
        using (var frame = e.OpenSkeletonFrame())
        {
            if (frame != null)
            {
                var skeletons = frame.Skeletons().Where(s => s.TrackingState == SkeletonTrackingState.Tracked);
                
                foreach (var skeleton in skeletons)
                {
                    if (skeleton != null)
                    {
                        // Update skeleton gestures.
                        gestureController.Update(skeleton);
                    }
                }
            }
        }
    }
    
    // ...
    
    void GestureController_GestureRecognized(object sender, GestureEventArgs e)
    {
        // Display the recognized gesture's name.
        Debug.WriteLine(e.Name);
    }
   

5. Recognizing and synthesizing voice:

    VoiceController voiceController = new VoiceController();
    voiceController.SpeechRecognized += VoiceController_SpeechRecognized;
    
    KinectSensor sensor = SensorExtensions.Default();
    List<string> phrases = new List<string> { "Hello", "Goodbye" };
    
    voiceController.StartRecognition(sensor, phrases);
    
    // ...
    
    void VoiceController_SpeechRecognized(object sender, Microsoft.Speech.Recognition.SpeechRecognizedEventArgs e)
    {
        string text = e.Result.Text;
        
        voiceController.Speak("I recognized the words: " + text);
    }
   

• Developed by Vangos Pterneas for LightBuzz
• Gesture detection partly based on Fizbin library, by Nicholas Pappas

You are free to use these libraries in personal and commercial projects by attributing the original creator of Vitruvius. Licensed under Apache v2 License.