Skeletal Tracking Fundamentals

Is it possible for kinect windows device to capture gestures of a person when sited down behind a desk? I understood that the currently available kinect device is not capable of partial skeleton detection. Is that true? from your sample video it seem that skeleton is clipped (that is a partial tracking), nevertheless you always stand up to be tracked.
thank you
mattia

I would like some help with this please. Does anyone have any source or information on the accuracy of the joint coordinates obtained by the Kinect Sensor? For example, whether the x and y coordinates have an error of +/- 0.05m and the z coordinate an error of +/- 0.07m (Just random values I put to help explain what I am looking for). Any help would be greatly appreciated. Thanks in advance!

Josiah

Hi! I have a question. In that sample code, scaledPosition method is not working.When I take a breakpoint at

Joint scaledJoint = joint.ScaleTo(1280, 720); 

in debug mode, scaledJointType is a HipCenter, not what I want. And trackingState fixed to 'NotTracked' mode.I don`t know what I have to do.please help. Thanks.

@Dan:Hmm. That`s strange. I didn`t modifiy the sample code. I re-downloaded it, but the same problem occured in that situation.Joint Information is right,but scaledjoint is still incorrect..

anyways. here is the code.

bool closing = false;
        const int skeletonCount = 6; 
        Skeleton[] allSkeletons = new Skeleton[skeletonCount];

        private void Window_Loaded(object sender, RoutedEventArgs e)
        {
            kinectSensorChooser1.KinectSensorChanged += new DependencyPropertyChangedEventHandler(kinectSensorChooser1_KinectSensorChanged);

        }

        void kinectSensorChooser1_KinectSensorChanged(object sender, DependencyPropertyChangedEventArgs e)
        {
            KinectSensor old = (KinectSensor)e.OldValue;

            StopKinect(old);

            KinectSensor sensor = (KinectSensor)e.NewValue;

            if (sensor == null)
            {
                return;
            }

            


            var parameters = new TransformSmoothParameters
            {
                Smoothing = 0.3f,
                Correction = 0.0f,
                Prediction = 0.0f,
                JitterRadius = 1.0f,
                MaxDeviationRadius = 0.5f
            };
            sensor.SkeletonStream.Enable(parameters);

            sensor.SkeletonStream.Enable();

            sensor.AllFramesReady += new EventHandler<AllFramesReadyEventArgs>(sensor_AllFramesReady);
            sensor.DepthStream.Enable(DepthImageFormat.Resolution640x480Fps30); 
            sensor.ColorStream.Enable(ColorImageFormat.RgbResolution640x480Fps30);

            try
            {
                sensor.Start();
            }
            catch (System.IO.IOException)
            {
                kinectSensorChooser1.AppConflictOccurred();
            }
        }

        void sensor_AllFramesReady(object sender, AllFramesReadyEventArgs e)
        {
            if (closing)
            {
                return;
            }

            //Get a skeleton
            Skeleton first =  GetFirstSkeleton(e);

            if (first == null)
            {
                return; 
            }



            //set scaled position
            //ScalePosition(headImage, first.Joints[JointType.Head]);
            ScalePosition(leftEllipse, first.Joints[JointType.HandLeft]);
            ScalePosition(rightEllipse, first.Joints[JointType.HandRight]);

            GetCameraPoint(first, e); 

        }

        void GetCameraPoint(Skeleton first, AllFramesReadyEventArgs e)
        {

            using (DepthImageFrame depth = e.OpenDepthImageFrame())
            {
                if (depth == null ||
                    kinectSensorChooser1.Kinect == null)
                {
                    return;
                }
                

                //Map a joint location to a point on the depth map
                //head
                DepthImagePoint headDepthPoint =
                    depth.MapFromSkeletonPoint(first.Joints[JointType.Head].Position);
                //left hand
                DepthImagePoint leftDepthPoint =
                    depth.MapFromSkeletonPoint(first.Joints[JointType.HandLeft].Position);
                //right hand
                DepthImagePoint rightDepthPoint =
                    depth.MapFromSkeletonPoint(first.Joints[JointType.HandRight].Position);


                //Map a depth point to a point on the color image
                //head
                ColorImagePoint headColorPoint =
                    depth.MapToColorImagePoint(headDepthPoint.X, headDepthPoint.Y,
                    ColorImageFormat.RgbResolution640x480Fps30);
                //left hand
                ColorImagePoint leftColorPoint =
                    depth.MapToColorImagePoint(leftDepthPoint.X, leftDepthPoint.Y,
                    ColorImageFormat.RgbResolution640x480Fps30);
                //right hand
                ColorImagePoint rightColorPoint =
                    depth.MapToColorImagePoint(rightDepthPoint.X, rightDepthPoint.Y,
                    ColorImageFormat.RgbResolution640x480Fps30);


                //Set location
                CameraPosition(headImage, headColorPoint);
                CameraPosition(leftEllipse, leftColorPoint);
                CameraPosition(rightEllipse, rightColorPoint);
            }        
        }


        Skeleton GetFirstSkeleton(AllFramesReadyEventArgs e)
        {
            using (SkeletonFrame skeletonFrameData = e.OpenSkeletonFrame())
            {
                if (skeletonFrameData == null)
                {
                    return null; 
                }

                
                skeletonFrameData.CopySkeletonDataTo(allSkeletons);

                //get the first tracked skeleton
                Skeleton first = (from s in allSkeletons
                                         where s.TrackingState == SkeletonTrackingState.Tracked
                                         select s).FirstOrDefault();

                return first;

            }
        }

        private void StopKinect(KinectSensor sensor)
        {
            if (sensor != null)
            {
                if (sensor.IsRunning)
                {
                    //stop sensor 
                    sensor.Stop();

                    //stop audio if not null
                    if (sensor.AudioSource != null)
                    {
                        sensor.AudioSource.Stop();
                    }


                }
            }
        }

        private void CameraPosition(FrameworkElement element, ColorImagePoint point)
        {
            //Divide by 2 for width and height so point is right in the middle 
            // instead of in top/left corner
            Canvas.SetLeft(element, point.X - element.Width / 2);
            Canvas.SetTop(element, point.Y - element.Height / 2);

        }

        private void ScalePosition(FrameworkElement element, Joint joint)
        {
            //convert the value to X/Y
            Joint scaledJoint = joint.ScaleTo(1280, 720); 
            
            //convert & scale (.3 = means 1/3 of joint distance)
            //Joint scaledJoint = joint.ScaleTo(1280, 720, .3f, .3f);

            Canvas.SetLeft(element, scaledJoint.Position.X);
            Canvas.SetTop(element, scaledJoint.Position.Y); 
            
        }


        private void Window_Closing(object sender, System.ComponentModel.CancelEventArgs e)
        {
            closing = true; 
            StopKinect(kinectSensorChooser1.Kinect); 
        }

only one that I modifiy is ScaledPosition Section.

Thanks for your help!

@Dan:Definitely. Can you explain why this happens? I can`t find the reason.. Thanks!  

I solve the problem. Thanks for your help!

So I followed your tutorial and cannot get the ScaleTo method to recognize for the Joint. Any idea why? I went and downloaded the sample code, but still cannot get it to work. What did you do myChan?

@Dan:

hi...i just wondering how doing the hover button with SDK V1?

refer link:

http://channel9.msdn.com/coding4fun/kinect/Recreating-the-Kinect-Hub#c634646980102097278

anyone knows?

need help. thanks

@Blas: Maybe I think, You should check the GetCameraPoint() method. To scale the position of skeleton, CameraPosition() method must be commented. And After you uncomment the ScaledPosition() method in sensor_AllFrameReady event, You can get a right result!

hi!

@myChan:

thanks! problem solved.

hmm...but how about the hover a button to select/click a target?

I tried that but nothing. I have your code exactly, its just the one line in line 180 where you call the joint.ScaleTo(1280, 720) method. It isn't recognizing the ScaleTo method even though I just downloaded this: http://c4fkinect.codeplex.com/ But its still not working for me. Where is the ScaleTo method in Joint?

@Blas: I think ScaleTo() is the built-in function of the joint.After you uncomment the line 180, you should comment the CameraPosition() (line 118~121),then you can get a right answer. Maybe CameraPosition() has a high priority than ScaledPosition(), I think..

@kendrick0772: Maybe You can get a inspration in here. Check this out.

http://social.msdn.microsoft.com/forums/en-US/wpf/thread/f2dbc76e-dd54-4be6-b7d8-a72b1f1296ac/

@myChan Well I can't do that...like I said I get an error with the scaleTo Method that it does not exist. Where is it? Where can I find it? Why does that error not show up in his code?

Never mind it works fine now!

Hello. I have problems with speed. You can find all details at http://social.msdn.microsoft.com/Forums/en-US/kinectsdk/thread/fe9df13c-94ba-4c4e-805e-51d4b33783fd

Do you have a solution?

Are there any sample code in C++ to get the X, Y, Z coordinate of the joints?

If I wanted to add images to the Camera and interact with them, how would I be able to detect if my either hands are touching the image? I tried using the PointToScreen method but I do not think this will work. Any ideas?

Hello Dan,
I have the same question as Mattia De Rosa.
"Is it possible for kinect windows device to capture gestures of a person when sited down behind a desk? I understood that the currently available kinect device is not capable of partial skeleton detection. Is that true? from your sample video it seem that skeleton is clipped (that is a partial tracking), nevertheless you always stand up to be tracked.
thank you
mattia"

Just to summarize is it possible to capture the hand gestures of a person seated behind a desk?

Thanks,
Vijay

@myChan:

how are u?

thanks for the reply!!! wonderful!

hmmm, i got questions to ask!

in the latest sdk v1.

how to make the cursor move no matter u are using left/right hand?

any sample code for that?

anyone knows?

thanks in advance!

Thanks for the tutorial!
Using sensor.SkeletonStream.Enable() (no 'smoothing' parameters then) I still have a significative delay. Can I further improve the response?
Or maybe it is a problem with my not-so-efficient laptop?

Thanks in advance ;)

Hi Dan,
Thanks a lot for your support with these tutorials, they are helpful. I trying to generate an aplication but I need to detect the pixel color, do you know how to do that?.
Thanks,
Regards.

Hi Dan,
For my application I would really like to be able to scale the joint distance but I can't get the ScaleTo method to recognize for the Joint.
It's this line I'm having trouble with:
Joint scaledJoint = joint.ScaleTo(1280, 720);

ScaleTo is not recognized. Is it because I'm missing a library or something of the sort?
Thanks,
Cheers!
Paola

Hi Dan,

Is it possible to limit the tracking to just one skeleton, so that when two or more skeletons appear in front of the kinect, only the skeleton which is recognized first is continued to be tracked? Thanks!

Hello Dan

Is it possible to get the skeleton coordinates given just a depth image. There is no Kinect involved.

Hey help plz :) how to get how many skeletons count in frame

Sir this tutotial is really very helpful, what i wanted to ask is that how we can perform the button click function using the Hand Gesture.
Plz guide me about it....

Can you plz upload a tutorial on how to use this control in an app?

Hi,
Does anyone know how what is the maximum distance a person could be from kinect and still be picked by skeleton tracking?

@Pramod: Find the detailed answer on the dept video =)
If I remember well it should be 1 to 4 meters for the xbox 360 kinect (default mode)
and 0.5 meters to 3 meters for the kinect for windows  who has near mode enable.

Check the video out

@Myra: Try using Ray Chambers video to guide you in the process. It helped me.
http://raychambers.wordpress.com/2012/04/04/kinect-mix-and-match/

 Hi @Dan 
  Have you ever tried making this  app using navigation windows in wpf? I tried but for some reason it doesn't work.

@ Dan

Hallo Dan,
i have an important question. I'm a student, and we work this time on a Kinect projekt. Your Videos are great and give a good introduction in programing with kinect.

I have a problem in the Skeleton Tracking Projekt. If switch the ColorViewer to autosize and maximize the Window, the Elipse on Hands are not longer on the right Position. Can you tell me how i make my app resizable?

SKELETON TRACKING PROBLEM

Hi, I've tried the demo you provide, and I've got problems with the 'skelate tracking". actually the exemple is not working well, I've even tried to copy the complete code, but it still not working. and by checking variable using brake point, ( first, allSkeletons, ...) it appears skeletons are not tracked.

Does anyone no the reason why or has n idea ?

Thank you in advance

near

SOLVED

  Oops Never mind for my question... it's solved ....
  Seems that I keep forgeting to add the windows loaded even in the window..

Hello

I want to track Hip Joints for 6 skeletons. Can anyone help with the code?

Hey, I need a bit of help here. Bear with me, I'm very new to C#. The video kind of jumps straight to the code. I just have a problem setting the ellipses. They don't move. I don't know how to bind them properly. How do you get the ellipses to move after you create them??

@Dan

Thanks a lot for that awesome Tutorial. I have been trying to apply it in my project but i kept receiving this warning which says "Warning: An ImageFrame instance was not Disposed." I am sure that i disposed all the frames that I created. Can you please help ?!

Here is my code:


using System;
using System.Collections.Generic;
using System.Linq;
using System.Text;
using System.Windows;
using System.Windows.Controls;
using System.Windows.Data;
using System.Windows.Documents;
using System.Windows.Input;
using System.Windows.Media;
using System.Windows.Media.Imaging;
using System.Windows.Navigation;
using System.Windows.Shapes;
using Microsoft.Kinect;
using Coding4Fun.Kinect.Wpf;
using System.Diagnostics;
using System.IO;

namespace KinectSkeleton
{
/// <summary>
/// Interaction logic for MainWindow.xaml
/// </summary>
public partial class MainWindow : Window
{
public MainWindow()
{
InitializeComponent();
}

bool closing = false;
const int skeletonCount = 6;
Skeleton[] allSkeletons = new Skeleton[skeletonCount];

private void Window_Loaded(object sender, RoutedEventArgs e)
{
myKinectSensorChooser.KinectSensorChanged += new DependencyPropertyChangedEventHandler(myKinectSensorChooser_KinectSensorChanged);
}

void myKinectSensorChooser_KinectSensorChanged(object sender, DependencyPropertyChangedEventArgs e)
{
KinectSensor oldSensor = (KinectSensor)e.OldValue;
if (oldSensor != null)
{
oldSensor.Stop();
oldSensor.AudioSource.Stop();
}

KinectSensor mySensor = (KinectSensor)e.NewValue;
if (mySensor == null)
return;

mySensor.DepthStream.Enable(DepthImageFormat.Resolution320x240Fps30);
mySensor.ColorStream.Enable();
mySensor.SkeletonStream.Enable();
mySensor.AllFramesReady += new EventHandler<AllFramesReadyEventArgs>(mySensor_AllFramesReady);

try
{
mySensor.Start();
Debug.WriteLine("Starting Sensor .....");
Debug.WriteLine("The Current Elevation Angle is: " + mySensor.ElevationAngle.ToString());
mySensor.ElevationAngle = 0;
}
catch (System.IO.IOException)
{
//another app is using Kinect
myKinectSensorChooser.AppConflictOccurred();
}
}

void mySensor_AllFramesReady(object sender, AllFramesReadyEventArgs e)
{
if (closing)
return;

byte[] depthImagePixels;
DepthImageFrame depthFrame = e.OpenDepthImageFrame();
if (depthFrame == null)
return;

depthImagePixels = GenerateDepthImage(depthFrame);
int stride = depthFrame.Width*4;
image1.Source =
BitmapSource.Create(depthFrame.Width, depthFrame.Height,
96, 96, PixelFormats.Bgr32, null, depthImagePixels, stride);


//Get a skeleton
Skeleton first = GetFirstSkeleton(e);
if (first == null)
return;

Debug.WriteLine("Head Position is : " + first.Joints[JointType.Head].ToString());

depthFrame.Dispose();

}

private byte[] GenerateDepthImage(DepthImageFrame depthFrame)
{
//get the raw data from the frame with the depth for every pixel
short[] rawDepthData = new short[depthFrame.PixelDataLength];
depthFrame.CopyPixelDataTo(rawDepthData);

//use frame to create the image to display on-screen
//frame contains color information for all pixels in image
//Height x Width x 4 (Red, Green, Blue, empty byte)
Byte[] pixels = new byte[depthFrame.Height * depthFrame.Width * 4];

//hardcoded locations to Blue, Green, Red (BGR) index positions
const int BlueIndex = 0;
const int GreenIndex = 1;
const int RedIndex = 2;
int player, depth;

//loop through all distances
//pick a RGB color based on distance
for (int depthIndex = 0, colorIndex = 0;
depthIndex < rawDepthData.Length && colorIndex < pixels.Length;
depthIndex++, colorIndex += 4)
{
//get the player (requires skeleton tracking enabled for values)
player = rawDepthData[depthIndex] & DepthImageFrame.PlayerIndexBitmask;

//gets the depth value
depth = rawDepthData[depthIndex] >> DepthImageFrame.PlayerIndexBitmaskWidth;

if (player > 0)
{
pixels[colorIndex + BlueIndex] = Colors.Gold.B;
pixels[colorIndex + GreenIndex] = Colors.Gold.G;
pixels[colorIndex + RedIndex] = Colors.Gold.R;
}
else
{
pixels[colorIndex + BlueIndex] = Colors.Green.B;
pixels[colorIndex + GreenIndex] = Colors.Green.G;
pixels[colorIndex + RedIndex] = Colors.Green.R;
}
}

return pixels;
}

Skeleton GetFirstSkeleton(AllFramesReadyEventArgs e)
{
using (SkeletonFrame skeletonFrameData = e.OpenSkeletonFrame())
{
if (skeletonFrameData == null)
{
return null;
}
skeletonFrameData.CopySkeletonDataTo(allSkeletons);

//get the first tracked skeleton
Skeleton first = (from s in allSkeletons
where s.TrackingState == SkeletonTrackingState.Tracked
select s).FirstOrDefault();

return first;
}
}

void StopKinect(KinectSensor sensor)
{
if (sensor != null)
{
if (sensor.IsRunning)
{
sensor.ElevationAngle = 0;
sensor.Stop();
if (sensor.AudioSource != null)
{
sensor.AudioSource.Stop();
}
}
}
}
private void Window_Closing(object sender, System.ComponentModel.CancelEventArgs e)
{
closing = true;
Debug.WriteLine("Closing window...");
StopKinect(myKinectSensorChooser.Kinect);
}
}
}

Hi,@Dan.
I am a student from Tsing Hua University, China. I've been following you tutorial these days. The first four sections went great. However, when I came to the fifth one, which is the Skeletal Tracking Fundamentals, a problem emerged.
I believe I had exactly followed your steps and codes showed in the video, but it just didn't work out. I mean the Image control and the two ellipses I had set on the MainWindow didn't follow my joints, that is my head and two hands as you have specified in the video.
Here is my code as you may want to check it out.
bool closing=false;
const int skeletonCount=6;
Skeleton[] allSkeletons=new Skeleton[skeletonCount];


private void Window_Loaded(object sender, RoutedEventArgs e)
{
kinectSensorChooser1.KinectSensorChanged += new DependencyPropertyChangedEventHandler(kinectSensorChooser1_KinectSensorChanged);
}

void kinectSensorChooser1_KinectSensorChanged(object sender, DependencyPropertyChangedEventArgs e)
{
KinectSensor oldSensor = (KinectSensor)e.OldValue;
StopKinect(oldSensor);
KinectSensor newSensor = (KinectSensor)e.NewValue;
newSensor.ColorStream.Enable();
newSensor.DepthStream.Enable();
newSensor.SkeletonStream.Enable();
newSensor.AllFramesReady += new EventHandler<AllFramesReadyEventArgs>(newSensor_AllFramesReady);
try
{
newSensor.Start();
}
catch (System.IO.IOException)
{

kinectSensorChooser1.AppConflictOccurred();

}


}

void newSensor_AllFramesReady(object sender, AllFramesReadyEventArgs e)
{

if(closing)
{
return;
}
Skeleton first=GetFirstSkeleton(e);
if(first==null)
{
return;
}
GetCameraPoint(first,e);



}
void GetCameraPoint(Skeleton first,AllFramesReadyEventArgs e)
{
using(DepthImageFrame depth=e.OpenDepthImageFrame())
{
if(depth==null||kinectSensorChooser1.Kinect==null)
{
return;
}
DepthImagePoint headDepthPoint=
depth.MapFromSkeletonPoint(first.Joints[JointType.Head].Position);
DepthImagePoint leftDepthPoint=
depth.MapFromSkeletonPoint(first.Joints[JointType.HandLeft].Position);
DepthImagePoint rightDepthPoint=
depth.MapFromSkeletonPoint(first.Joints[JointType.HandRight].Position);
ColorImagePoint headColorPoint=
depth.MapToColorImagePoint(headDepthPoint.X,headDepthPoint.Y,
ColorImageFormat.RgbResolution640x480Fps30);
ColorImagePoint leftColorPoint=
depth.MapToColorImagePoint(leftDepthPoint.X,leftDepthPoint.Y,
ColorImageFormat.RgbResolution640x480Fps30);
ColorImagePoint rightColorPoint=
depth.MapToColorImagePoint(rightDepthPoint.X,rightDepthPoint.Y,
ColorImageFormat.RgbResolution640x480Fps30);


CameraPosition(headImage, headColorPoint);
CameraPosition(leftellipse, leftColorPoint);
CameraPosition(rightellipse, rightColorPoint);

}


}
private void CameraPosition(FrameworkElement element, ColorImagePoint point)
{
Canvas.SetLeft(element,point.X-element.Width/2);
Canvas.SetTop(element,point.Y-element.Height/2);

}

Skeleton GetFirstSkeleton(AllFramesReadyEventArgs e)
{
using(SkeletonFrame skeletonFrameData =e.OpenSkeletonFrame())
{
if(skeletonFrameData==null)
{
return null;
}
skeletonFrameData.CopySkeletonDataTo(allSkeletons);
Skeleton first=(from s in allSkeletons
where s.TrackingState==SkeletonTrackingState.Tracked
select s).FirstOrDefault();
return first;
}


}

void StopKinect(KinectSensor sensor)
{
if(sensor!=null)
{
sensor.Stop();
sensor.AudioSource.Stop();


}

}

private void Window_Closing(object sender, System.ComponentModel.CancelEventArgs e)
{
StopKinect(kinectSensorChooser1 .Kinect);


}

private void SetTilt_Click(object sender, RoutedEventArgs e)
{
SetTilt.IsEnabled = false;
if (kinectSensorChooser1.Kinect != null && kinectSensorChooser1.Kinect.IsRunning)
{
kinectSensorChooser1.Kinect.ElevationAngle = (int)slider1.Value;

}
System.Threading.Thread.Sleep(new TimeSpan(hours:0,minutes:0,seconds:1));
SetTilt.IsEnabled = true;

}

private void slider1_ValueChanged(object sender, RoutedPropertyChangedEventArgs<double> e)
{

}
I was wondering if you could help me a little bit. And I would be so grateful.

@Adila & @kinjad: I had a similar problem and it was resolved by removing all 'HorizontalAlignment' and 'VerticalAlignment' properties from the elements you are wanting to move.

Hope that helps,

Hi Dan..

When i try to run my code, which basically tracks my right hand, and ellipses appear on my shoulder, hand and elbow joints, i get an error like this, while running:

"NullReferenceException was unhandled". , on this line of code:

Skeleton skeleton= (from s in allSkeletons where s.TrackingState == SkeletonTrackingState.Tracked select s).FirstOrDefault(); 

The Bold part on the above line is where the error is highlighted in my code.

Can you help me with this, coz i'm kinda stuck with this for days now..

Thanks in Advance.. 

hi,
can any1 mail me or upload the code(c++ in opencv) for gesture recognition using kinect.
here is mail id
vikas.mulage@gmail.com

@Vikas Meet me tom

Hi Dan...,
I just went through tutorial its awesome!!,
i'm just working on Real time Motion Retargeting into 3D model using Kinect,i'm unable to proceed after skeleton data extraction,please help me out regarding further steps
Thanks,
Regards>

Hi Dan...,

It would be possible with the tool box do a mob cauter or a crwd couter?

associate face images to a square?

Would not use the skeleton, but as you associate the image to the face .. and an algorithm to count the squares!

Can you help me implement this?

something like this video:

http://www.youtube.com/watch?v=RYuDiQDM0MM&feature=related

thanx

Art

Hello Dan,

I am trying to control a Drone with the kinect. Do i use the skeleton and depth data for that?

I am running out of time . please help.

Hi Dan,

The sample code runs perfectly on my computer; however, when exiting the application, an exception happened on line 177 of KinectSkeletonViewer.xaml.cs where the line reads "ColorImagePoint colorPoint = depthFrame.MapToColorImagePoint(depthPoint.X, depthPoint.Y, this.Kinect.ColorStream.Format);".

The warning message is shown as "InvalidOperationException was unhandled
This API has returned an exception from an HRESULT: 0x80070015".

Hi Dan,

Awesome! Thanks for the video.

I've try out the skeleton tracking (full body) and it work well for the standing/vertical position. However, when I try to do a bit of the exercise in which requires lying down and the skeleton in more of a horizontal position the tracking does not work well. May I check with you if the there is anyway to work around this problem ? or would there be a release in the future of the sdk that would incorporate this lying down/horizontal skeleton tracking features.

Please help.

Thanks,

Viet 

@Dan: but I go here http://www.microsoft.com/en-us/kinectforwindows/Develop/Developer-Downloads.aspx and the SDK is still 1.5, the toolkit is the one that have been updated. I'm getting the same "InvalidOperationException was unhandled This API has returned an exception from an HRESULT: 0x80070015" and don't really know what to do, cause it's related to the SkeletonViewer.

If I delete that viewer from the mainWindow.xaml, that message doesn't appear (not even at closing the window), but that means that I cannot render the Skeletons on Screen automatically.

Any ideas? Thanks.

Hello!

I have a question.
I've built the demo like that above. But when I've put ellipses for joints on canvas and run the project, ellipses did't appeared in front of color picture. Instead they are in background - not visible. What I did wrong?

Thank you for the answer.

@Dan: I am trying to use the design editor for the main window of the Visual Basic code sample, but i am getting Invalid Markup error. No changes to original code was made. Using KSDK 1.0

THanks!

@Perry and @juanpibanez

I have a solution to the "InvalidOperationException was unhandled This API has returned an exception from an HRESULT: 0x80070015" and I figured I would post for any future readers.

I installed version 1.6 of the SDK and the Toolkit (Kinect for Windows) to no avail. After this, I opened the KinectSkeletonViewer.xaml.cs and noticed some function were now obsolete. To make the fix, change the following:

Line 172 - 181
//******************
CoordinateMapper cm = new CoordinateMapper(this.Kinect);
DepthImagePoint depthPoint = cm.MapSkeletonPointToDepthPoint(skeletonPoint, depthFrame.Format);

switch (ImageType)
{
case ImageType.Color:
ColorImagePoint colorPoint = new ColorImagePoint();
if (this.Kinect != null && this.Kinect.IsRunning)
colorPoint = cm.MapDepthPointToColorPoint(depthFrame.Format, depthPoint, this.Kinect.ColorStream.Format);

// map back to skeleton.Width & skeleton.Height
//******************

Once I updated these operations to the revised functions, I no longer receive the error. Hope this helps!

I used the QuickStart sample code for skeletal tracking and have the latest SDK, but I dont understand why the headImage does not follow my movements. The ellipses follow my hand motions. Why is it when I run it, I can get the ellipses to work, but the headImage automatically goes to the top left corner of the window and doesn't follow my head motions?

I was partially able to solve the problem. All 3 joints in the code are now tracking and objects moving, but I changed the headImage image object to an ellipse object in the .xaml window. The question now is why the headImage image object did not work but the ellipse object works.

I really like your videos, they have been very helpful to me. I want to save a movement (x, y, z) and then compare this with another movement (x, y, z) in real time, the idea is that the user will try to mimic the movement saved, and the system makes a calculation to find out how similar (0% ~ 100%) the motions are. My current problem is that the skeleton  from the recorded movement is different of the skeleton that is trying to mimic the movement, I'm trying to work with the idea of ​​a standardized skeleton you have any suggestions on how I can record and compare data in a standardized way?

Hi,

I tried to follow the video. Everything compiles, but the ellipses and image are not well mapped in the window. I checked CameraPosition and I do have the same calculation for the "SetLeft" and "SetTop".

Also, I am unable to download the sample code.

Thanks for helping

The download link of the source files is dead:(

Can you fix this?

DL Link: http://adf.ly/EP7dk

@SkiRacerDude Thank you:)

Hi Channel 9, this is really a beautiful series that you put up.

Hi I was just wondering if you have your videos transcribed somewhere for download? I really appreciate the tips I'm learning but English is not my native language, so sometimes the presenter speak a bit too fast for me to follow. Would really appreciate a text version of these tutorials.

Keep up the good work. Thank you very much.

@Basvm: Welcome

Dear Sir
Im working on kinect, i have some questions
1- hip center, hip_left and hip_right . what does represent exactly in the body, hip its not a point its a part consist of 3 items in the body,
please tell me what excatly represent in the body?

2- for depth data, it gives only distance between object and kinect, if i have object 3d like cube , for points in the z-axis it also retuen this points or just points in front of kinect?

3- if i need to track points that doesnt exist in the available joints , can i add other joints/points or not?

Warning 1 'Microsoft.Kinect.DepthImageFrame.MapFromSkeletonPoint(Microsoft.Kinect.SkeletonPoint)' is obsolete: 'This method is replaced by Microsoft.Kinect.CoordinateMapper.MapSkeletonPointToDepthPoint'

depth.MapFromSkeletonPoint(first.Joints[JointType.Head].Position);

this is the line that it say it obsolete.

anyone know how to fix this problem.

Hi,
trying to run the sample code and came up with a problem :
the line "using Microsoft.Kinect;" has a red line under Kinect. the program is runing, but a build error is showing before. anyone knows how can I make this red line under the Kinect to disapear?

hiiii

how can i do when Hover button by hand skeleton and the command load image to tracking spine skeleton in c#?

thanksssssss

Thanks for great tutorial.
One thing: how to set the image with transparent background? I tried to do so, but the backgroud is always white.

Hi Dan,

Really helpful tutorial. Helps me a lot.
But, I am stuck in a very basic problem. How can I record RGB and depth video. I tried to save RGB and depth as .png files (as shown in colorBasics example, writing one .png color image) in each event handler method. But, it generates 6fps. How can I record RGB and depth video in 30fps speed? Do I need to use multi-threading? I don't know how to use it. Your help regarding this will be highly appreciated and very helpful for the beginners.

Thanks again.

Also, the RGB and depth images are not synchronous when I tried to write in this way. Means, for all RGB images, there may not be a similar depth image.

@Adam Victor See my post above. I believe that is the fix you are looking for.

Hi dan.

Thank you for all the tutorials, are very useful. I´m doing an school Project and i need to recognize some patterns like rectangles, circles and then know at what depht they are. Do you have some example of this or somewhere can i find something like this?. Thank you again for your tutorials, keep working like that.

I am developing a kinect app which needs to be able to save the join positions at set intervals , depth won't matter though so I only need to store the x and y I thin , My initial Idea was to do this through image comparison but I want to use the skeleton data , any help much appreciated

Please, Could you update this sample for the latest SDK.. It's really great and it's helping a lot in my graduation project, but now it's giving me many warnings with new SDK. Could you please help?

Error 9 The name 'headImage' does not exist in the current

I am gettimg the same error for "leftEllipse" and "rightEllipse". Does anyone can help me debug this issue ?...

Any help is appreciated

Hi Dan,

I have a strange error with these constructor functions...

GetFirstSkeleton(e);
ScalePosition
GetCameraPoint(first, e);

All the same error "The Name dose not exist for the current context"

Hi Dan

The thing I do to the sample is to change the reference Microsoft.Kinect to a new version 1.7.0.0.Then I change KinectSkeletonViewer.xaml.cs as Mason told before.But I still get something problems.

Firstly, when I close the MainWindow by clicking the X at the righttop cornor,the image stop moving. However, the MainWindow is still there and can't be closed. In Windows Task Manager,it shows that the MainWindow is not responding. Only after I stop debugging, the window will be closed. So what's wrong with this? And what should I do?


Secondly, I get 6 warnings related to obsolete in line 128 131 134 140 144 148. For example:
'Microsoft.Kinect.DepthImageFrame.MapFromSkeletonPoint(Microsoft.Kinect.SkeletonPoint)' is obsolete: '"This method is replaced by Microsoft.Kinect.CoordinateMapper.MapSkeletonPointToDepthPoint"
I am new to C# so I don't know how to change that.However the program still goes well with the warning.