（四）Kinect人脸识别

kinect可以通过摄动摄像头不仅可以获取人脸位置旋转信息，也可以获取脸部轮廓的三维坐标

可以参考插件中的场景KinectFaceTrackingDemo1-4，在kinectManager基础上需要脚本FacetrackingManager。

1）通过KinectManager kinectManager = KinectManager.Instance;获取到Kinect运行实例

2）通过sensorData = kinectManager.GetSensorData();获取到传感器信息

3）同SensorData类的sensorInterface获取相关信息，如下为部分方法，可以获得人脸的Rect以及头部位置等等

 

    bool IsFaceTrackingAvailable(ref bool bNeedRestart);

    // initializes libraries and resources needed by the face tracking subsystem
    bool InitFaceTracking(bool bUseFaceModel, bool bDrawFaceRect);

    // releases the resources and libraries used by the face tracking subsystem
    void FinishFaceTracking();

    // this method gets invoked periodically to update the face tracking state
    // returns true if update is successful, false otherwise
    bool UpdateFaceTracking();

    // returns true if face tracking is initialized, false otherwise
    bool IsFaceTrackingActive();

    // returns true if face rectangle(s) must be drawn in color map, false otherwise
    bool IsDrawFaceRect();

    // returns true if the face of the specified user is being tracked at the moment, false otherwise
    bool IsFaceTracked(long userId);

    // gets the face rectangle in color coordinates. returns true on success, false otherwise
    bool GetFaceRect(long userId, ref Rect faceRect);

    // visualizes face tracker debug information
    void VisualizeFaceTrackerOnColorTex(Texture2D texColor);

    // gets the head position of the specified user. returns true on success, false otherwise
    bool GetHeadPosition(long userId, ref Vector3 headPos);
    
    // gets the head rotation of the specified user. returns true on success, false otherwise
    bool GetHeadRotation(long userId, ref Quaternion headRot);