使用Kinect获取面部点的深度信息

时间:2012-11-22 16:13:04

标签: c# sdk kinect face-recognition depth

我目前正在使用微软的SDK跟踪一张脸,并且可以检测到脸上的121点。我可以得到以下庄园中各点的坐标:

            if (!this.lastFaceTrackSucceeded || this.skeletonTrackingState != SkeletonTrackingState.Tracked)
                return;

            var faceModelPts = new List<Point>();
            var faceModel = new List<FaceModelTriangle>();

            for (int i = 0; i < this.facePoints.Count; i++)
            {
                faceModelPts.Add(new Point(this.facePoints[i].X + 0.5f, this.facePoints[i].Y + 0.5f));
                FaceDataPoints.XPointInfo[i] = this.facePoints[i].X;
                FaceDataPoints.YPointInfo[i] = this.facePoints[i].Y;
            }

但这些点似乎是彩色图像的坐标。如何获取有关每个像素的相应深度信息。比如用户脸上点“20”?

1 个答案:

答案 0 :(得分:0)

找到答案

“GetProjected3DShape”形状保存面部跟踪基础使用的2D坐标

要获取深度信息,请使用“Get3DShape”

比较2

        private EnumIndexableCollection<FeaturePoint, PointF> ColourPoints;

        private EnumIndexableCollection<FeaturePoint, Vector3DF> DepthPoints;

深度点保存每个面部点的X,Y和Z坐标信息