Kinect for windows的脸部识别
原文: http://axzxs.blog.51cto.com/730810/1191261
需要引入的dll:http://down.51cto.com/data/774502
Kinect for windows提供了脸部识别的功能,可以识出人脸。主要是通过FaceTrackFrame类的GetTriangles()来得到一个三角形数组,这个三角形数组就是给成人面部的基本形状,并且组成的效果是立体的(可以这样理解,可以把3D都拆成三角形来表示,看上去像3D,但其实是2D),这个数组的每个元素都存放着三个整数,分别代码三角形的第一个点,第二个点和第三个点。FaceTrackFrame的GetProjected3DShape方法,可以获取一组坐标信息,这样就可以结合三角形数组元素中的点作为索引,从本方法的坐标集合中取出每个三角形的坐标点来了,就可以绘制这些三角形,就可以组成一个人脸的网络3D效果图了。
本例是从色彩摄像头中获取彩色数据流,并显示到窗体上,再通过FaceTracker得到得到脸部3D信息,并用GDI+的方式把网络图形画到窗体上,这时就可以在真实图像上看到浮着一张网络的面套。同时可以得出脸部交汇最多的坐标,并用GDI+添加上不同有色彩,同时还可以得到人体面部和Kinect 正面的偏差,即人头是否竖直,有没有偏上一边等角度信息。
using System; using System.Collections.Generic; using System.ComponentModel; using System.Data; using System.Drawing; using System.Linq; using System.Text; using System.Threading.Tasks; using System.Windows.Forms; using Microsoft.Kinect; using Microsoft.Kinect.Toolkit; using Microsoft.Kinect.Toolkit.FaceTracking; using System.Threading; using System.IO; using System.Drawing.Imaging; namespace Face { public partial class Form1 : Form { public Form1() { InitializeComponent(); } KinectSensor ks = null; private void Form1_Load(object sender, EventArgs e) { //让winform窗体刷新不闪动 this.SetStyle(ControlStyles.OptimizedDoubleBuffer, true); this.SetStyle(ControlStyles.AllPaintingInWmPaint, true); this.SetStyle(ControlStyles.UserPaint, true); this.SetStyle(ControlStyles.DoubleBuffer, true); //找到连接的Kinect设备 foreach (var ks in KinectSensor.KinectSensors) { if (ks.Status == KinectStatus.Connected) { this.ks = ks; } } //开启色彩流,深度流,骨骼流的跟踪 if (this.ks != null) { this.ks.ColorStream.Enable(ColorImageFormat.RgbResolution640x480Fps30); this.ks.DepthStream.Enable(DepthImageFormat.Resolution320x240Fps30); this.ks.DepthStream.Range = DepthRange.Near; this.ks.SkeletonStream.EnableTrackingInNearRange = true; this.ks.SkeletonStream.TrackingMode = SkeletonTrackingMode.Seated; this.ks.SkeletonStream.Enable(); //订阅跟踪数据读取事件 this.ks.AllFramesReady += OnAllFramesReady; ks.Start(); } } //这个方法很重要,就是重绘人脸跟踪采集到的数据 protected override void OnPaint(PaintEventArgs e) { base.OnPaint(e); foreach (SkeletonFaceTracker faceInformation in this.trackedSkeletons.Values) { //第一个参数为当前窗体为画布,第二个是添加采集到的信息到listbox中,这个方法画识别到脸部的信息 faceInformation.DrawFaceModel(e.Graphics, Messbox_LB); } } //定义脸部识别的集合 private readonly Dictionary<int, SkeletonFaceTracker> trackedSkeletons = new Dictionary<int, SkeletonFaceTracker>(); //色彩流字节数组 private byte[] colorImage; private ColorImageFormat colorImageFormat = ColorImageFormat.Undefined; //深度流字节数组 private short[] depthImage; private DepthImageFormat depthImageFormat = DepthImageFormat.Undefined; //骨骼信息数组 private Skeleton[] skeletonData; private void OnAllFramesReady(object sender, AllFramesReadyEventArgs allFramesReadyEventArgs) { ColorImageFrame colorImageFrame = null; DepthImageFrame depthImageFrame = null; SkeletonFrame skeletonFrame = null; try { colorImageFrame = allFramesReadyEventArgs.OpenColorImageFrame(); //接到色彩流对框架 depthImageFrame = allFramesReadyEventArgs.OpenDepthImageFrame(); //接到深度流对框架 skeletonFrame = allFramesReadyEventArgs.OpenSkeletonFrame(); //接到骨骼流对框架 if (colorImageFrame == null || depthImageFrame == null || skeletonFrame == null) { return; } if (this.depthImageFormat != depthImageFrame.Format) { this.ResetFaceTracking(); this.depthImage = null; this.depthImageFormat = depthImageFrame.Format; } if (this.colorImageFormat != colorImageFrame.Format) { this.ResetFaceTracking(); this.colorImage = null; this.colorImageFormat = colorImageFrame.Format; } if (this.depthImage == null) { this.depthImage = new short[depthImageFrame.PixelDataLength]; } if (this.colorImage == null) { this.colorImage = new byte[colorImageFrame.PixelDataLength]; } if (this.skeletonData == null || this.skeletonData.Length != skeletonFrame.SkeletonArrayLength) { this.skeletonData = new Skeleton[skeletonFrame.SkeletonArrayLength]; } //获取各种数据流信息 colorImageFrame.CopyPixelDataTo(this.colorImage); depthImageFrame.CopyPixelDataTo(this.depthImage); skeletonFrame.CopySkeletonDataTo(this.skeletonData); //清空列表信息 Messbox_LB.Items.Clear(); //编历骨骼流 foreach (Skeleton skeleton in this.skeletonData) { //找到有效的骨骼信息 if (skeleton.TrackingState == SkeletonTrackingState.Tracked || skeleton.TrackingState == SkeletonTrackingState.PositionOnly) { if (!this.trackedSkeletons.ContainsKey(skeleton.TrackingId)) { //添加骨骼信息到集合中 this.trackedSkeletons.Add(skeleton.TrackingId, new SkeletonFaceTracker()); } // 得到脸部识别对象 SkeletonFaceTracker skeletonFaceTracker; if (this.trackedSkeletons.TryGetValue(skeleton.TrackingId, out skeletonFaceTracker)) { //把获取的数据流的相关信息传给OnFrameReady方法 skeletonFaceTracker.OnFrameReady(Messbox_LB, this.ks, colorImageFormat, colorImage, depthImageFormat, depthImage, skeleton); skeletonFaceTracker.LastTrackedFrame = skeletonFrame.FrameNumber; } } } //这个刷新会触发窗体的重画,OnPaint方法会被调用。 this.Refresh(); //把色彩流转转成位图显示成窗体的背景 this.BackgroundImage = ToGrayBitmap(colorImage, 640, 480); } finally { if (colorImageFrame != null) { colorImageFrame.Dispose(); } if (depthImageFrame != null) { depthImageFrame.Dispose(); } if (skeletonFrame != null) { skeletonFrame.Dispose(); } } } //把色采流数据转成位图返回 public static Bitmap ToGrayBitmap(byte[] rawValues, int width, int height) { //定议转换图片的格式,一个像素占32个,前24位为红绿蓝,后8位为空 PixelFormat pf = PixelFormat.Format32bppRgb; //申请目标位图的变量 Bitmap bmp = new Bitmap(width, height, pf); //将其内存区域锁定 BitmapData bmpData = bmp.LockBits(new Rectangle(0, 0, width, height), ImageLockMode.WriteOnly, pf); //获取位图的起始地址 IntPtr iptr = bmpData.Scan0; //用Marshal的Copy方法,将色彩流字节数组复制到BitmapData中 System.Runtime.InteropServices.Marshal.Copy(rawValues, 0, iptr, rawValues.Length); //释放锁 bmp.UnlockBits(bmpData); return bmp; } //重新设置识别对象 private void ResetFaceTracking() { foreach (int trackingId in new List<int>(this.trackedSkeletons.Keys)) { this.RemoveTracker(trackingId); } } //从集合中移动识别信息 private void RemoveTracker(int trackingId) { this.trackedSkeletons[trackingId].Dispose(); this.trackedSkeletons.Remove(trackingId); } private void Form1_FormClosing(object sender, FormClosingEventArgs e) { if (this.ks.Status == KinectStatus.Connected) { ks.Stop(); } } //定义脸识别类 class SkeletonFaceTracker : IDisposable { //定义脸部识别形状三角形数组 private static FaceTriangle[] faceTriangles; //脸部识别坐标点集合 private EnumIndexableCollection<FeaturePoint, Microsoft.Kinect.Toolkit.FaceTracking.PointF> facePoints; //脸部跟踪类 private FaceTracker faceTracker; //定义识别成功标识 private bool lastFaceTrackSucceeded; //骨骼跟踪状态 private SkeletonTrackingState skeletonTrackingState; public int LastTrackedFrame { get; set; } public void Dispose() { if (this.faceTracker != null) { this.faceTracker.Dispose(); this.faceTracker = null; } } //用来把识别的信息绘制出来 public void DrawFaceModel(Graphics graphics, ListBox lb) { if (!this.lastFaceTrackSucceeded || this.skeletonTrackingState != SkeletonTrackingState.Tracked) { return; } List<System.Drawing.PointF> faceModelPts = new List<System.Drawing.PointF>(); for (int i = 0; i < this.facePoints.Count; i++) { faceModelPts.Add(new System.Drawing.PointF(this.facePoints[i].X + 0.5f, this.facePoints[i].Y + 0.5f)); } System.Drawing.Pen pen = new System.Drawing.Pen(System.Drawing.Color.Green); List<System.Drawing.PointF> list = new List<System.Drawing.PointF>(); //遍历所有的三角形,分别画三角形 for (int i = 0; i < faceTriangles.Count(); i++) { System.Drawing.PointF[] pointFarr = new System.Drawing.PointF[4]; pointFarr[0] = faceModelPts[faceTriangles[i].First]; pointFarr[1] = faceModelPts[faceTriangles[i].Second]; pointFarr[2] = faceModelPts[faceTriangles[i].Third]; pointFarr[3] = faceModelPts[faceTriangles[i].First]; list.AddRange(pointFarr.Take(3)); graphics.DrawLines(pen, pointFarr); } lb.Items.Add(list.GroupBy(f => f).Count() + "点"); int count = list.GroupBy(f => f).Max(s => s.Count()); lb.Items.Add(count); foreach (var v in list.GroupBy(f => f).Where(s => s.Count() == 10)) { lb.Items.Add(v.Key + " " + 10); graphics.FillEllipse(new SolidBrush(System.Drawing.Color.Red), v.Key.X, v.Key.Y, 5, 5); } foreach (var v in list.GroupBy(f => f).Where(s => s.Count() == 9)) { lb.Items.Add(v.Key + " " + 9); graphics.FillEllipse(new SolidBrush(System.Drawing.Color.Blue), v.Key.X, v.Key.Y, 5, 5); } foreach (var v in list.GroupBy(f => f).Where(s => s.Count() == 8)) { lb.Items.Add(v.Key + " " + 8); graphics.FillEllipse(new SolidBrush(System.Drawing.Color.Black), v.Key.X, v.Key.Y, 5, 5); } } /// <summary> /// 数据更新的方法 /// </summary> internal void OnFrameReady(ListBox lb, KinectSensor kinectSensor, ColorImageFormat colorImageFormat, byte[] colorImage, DepthImageFormat depthImageFormat, short[] depthImage, Skeleton skeletonOfInterest) { this.skeletonTrackingState = skeletonOfInterest.TrackingState; //判断是否为跟踪状态 if (this.skeletonTrackingState != SkeletonTrackingState.Tracked) { return; } if (this.faceTracker == null) { try { //从KinectSensor中实例化出一个脸部识别对象 this.faceTracker = new FaceTracker(kinectSensor); } catch (InvalidOperationException) { this.faceTracker = null; } } if (this.faceTracker != null) { //从脸部识别对象中得到脸识别框架 FaceTrackFrame frame = this.faceTracker.Track( colorImageFormat, colorImage, depthImageFormat, depthImage, skeletonOfInterest); //标识识别成功 this.lastFaceTrackSucceeded = frame.TrackSuccessful; if (this.lastFaceTrackSucceeded) { if (faceTriangles == null) { //得到脸部识别三角形数组 faceTriangles = frame.GetTriangles(); } //得到脸部识别点的坐标 this.facePoints = frame.GetProjected3DShape(); //加载脸部的空间位置 lb.Items.Add("Rotation 仰低头:" + frame.Rotation.X); lb.Items.Add("Rotation 左右转头:" + frame.Rotation.Y); lb.Items.Add("Rotation 左右偏头:" + frame.Rotation.Z); } } } } } } }