1

我正在做我的项目，即机器人空气曲棍球运动员；

在项目中，我使用 Microsoft lifecam 来识别“球”的位置。那行得通。

问题是当我以主窗体运行项目时 - 如果在进入游戏之前我正在检查相机（一张图片），它是按钮内的代码：

 Capture capture = new Capture(0);
 Image<Emgu.CV.Structure.Bgr,byte>  temp = capture.QueryFrame();
 picture pic = new picture(temp);
 pic.ShowDialog();

并且在我盯着玩游戏的检查之后停止给我这个错误：

An unhandled exception of type 'System.AccessViolationException' 
occurred in Emgu.CV.dll
Additional information: Attempted to read or write protected memory. 
This is often an indication that other memory is corrupt.

图像处理代码为：

 class C_Camera
    {
        public Capture capWebCam = null;
        public bool binCapturingInProgress = false;
        public Image<Bgr, byte> imgOriginal;
        public Image<Gray, byte> imgProcessed;
        public ImageBox imageBox1 = new ImageBox();
        public TextBox textBox1;
        public List<PointF> centerList;
        CircleF[] circles;

        public void newCapture(bool b, List<PointF> centers)
        {
            centerList = centers;
            binCapturingInProgress = b;

            try
            {
                capWebCam = new Capture();
            }
            catch (NullReferenceException except)
            {
                MessageBox.Show(except.Message);
                return;
            }

            Application.Idle += processFrameAndUpdateGUI;//add process image function to the applicationlist in tasks
            binCapturingInProgress = true;
        }


        public void processFrameAndUpdateGUI(object sender, EventArgs arg)
        {
            imgOriginal = capWebCam.QueryFrame();
            if (imgOriginal == null) 
            {
                return;
            }
            imgProcessed = imgOriginal.InRange(new Bgr(0, 0, 150)/*min filter*/, new Bgr(80, 80, 256));
         //   imgProcessed = imgOriginal.InRange(new Bgr(0, 0, 175)/*min filter*/, new Bgr(100, 100, 256));

            imgProcessed = imgProcessed.SmoothGaussian(9);//9
            circles = imgProcessed.HoughCircles(new Gray(100), //canny threshhold
                                                          new Gray(50), //accumolator threshold
                                                          2,//size of image 
                                                          imgProcessed.Height / 4, //min size in pixels between centers of detected circles
                                                          15, //min radios
                                                          43)[0];//max radios and get circles from first channel
            foreach (CircleF circle in circles)
            {


                centerList.Add(circle.Center);
                //     circlesOfBalls.Add(circle);

                if (textBox1.Text != "") textBox1.AppendText(Environment.NewLine);
                textBox1.AppendText("ball position = x" + centerList[centerList.Count - 1].X +
                                    ", y = " + centerList[centerList.Count - 1].Y.ToString().PadLeft(4)
                                    + "centerList.Count= " + centerList.Count);

                // + ", radios = " + centerList[centerList.Count - 1].Radius.ToString("###,000").PadLeft(7))

                textBox1.ScrollToCaret();// scrolls the textBox to last line

                // draw a small green circle in the center
                CvInvoke.cvCircle(imgOriginal, // draw on the original image
                                  new Point((int)circle.Center.X, (int)circle.Center.Y),//center point of circle
                                  3, // radios of circle,
                                  new MCvScalar (0, 255, 0), // draw in green color
                                  -1, //indicates to fill the circle
                                  LINE_TYPE.CV_AA, //smoothes the pixels
                                  0);// no shift 

                //draw a red circle around the detected object
              //  if (imgOriginal.Data != null) 
                    imgOriginal.Draw(circle,    //current circle
                                new Bgr(Color.Red), // draw pure red
                                3);
            } // end of for each

           // if(imgOriginal.Data != null)
                imageBox1.Image = imgOriginal;
        }

我试图在谷歌和这里搜索答案，但找不到任何答案。如果我没有进入相机检查它工作得很好。

4

1 回答 1

0

从外观上看，您只有在我看到您应该在系统空闲时捕获图像时才捕获图像，但这可能太频繁或不够频繁。

尝试添加 DispatcherTimer 并为其提供渲染优先级和 1000/60（每秒 60 帧）的频率。

    DispatcherTimer UpdateTimer; 
    public Capture capWebCam = null;
    public bool binCapturingInProgress = false;
    public Image<Bgr, byte> imgOriginal;
    public Image<Gray, byte> imgProcessed;
    public ImageBox imageBox1 = new ImageBox();
    public TextBox textBox1;
    public List<PointF> centerList;
    CircleF[] circles;

    public void newCapture(bool b, List<PointF> centers)
    {
        centerList = centers;
        binCapturingInProgress = b;

        try
        {
            capWebCam = new Capture();
        }
        catch (NullReferenceException except)
        {
            MessageBox.Show(except.Message);
            return;
        }

        binCapturingInProgress = true;


        //Start the timer after the device is ready to be captured (this looks like a good place to do it)
        UpdateTimer = new DispatcherTimer(DispatcherPriority.Render);
        UpdateTimer.Interval = TimeSpan.FromMilliseconds(1000 / 60);
        UpdateTimer.Tick += processFrameAndUpdateGUI;
        UpdateTimer.Start();
    }

    private void processFrameAndUpdateGUI(object o, EventArgs e)
    {
       //Capture new image, render it and display it.

        imgOriginal = capWebCam.QueryFrame();
        if (imgOriginal == null) 
        {
            return;
        }
        imgProcessed = imgOriginal.InRange(new Bgr(0, 0, 150)/*min filter*/, new Bgr(80, 80, 256));
     //   imgProcessed = imgOriginal.InRange(new Bgr(0, 0, 175)/*min filter*/, new Bgr(100, 100, 256));

        imgProcessed = imgProcessed.SmoothGaussian(9);//9
        circles = imgProcessed.HoughCircles(new Gray(100), //canny threshhold
                                                      new Gray(50), //accumolator threshold
                                                      2,//size of image 
                                                      imgProcessed.Height / 4, //min size in pixels between centers of detected circles
                                                      15, //min radios
                                                      43)[0];//max radios and get circles from first channel
        foreach (CircleF circle in circles)
        {


            centerList.Add(circle.Center);
            //     circlesOfBalls.Add(circle);

            if (textBox1.Text != "") textBox1.AppendText(Environment.NewLine);
            textBox1.AppendText("ball position = x" + centerList[centerList.Count - 1].X +
                                ", y = " + centerList[centerList.Count - 1].Y.ToString().PadLeft(4)
                                + "centerList.Count= " + centerList.Count);

            // + ", radios = " + centerList[centerList.Count - 1].Radius.ToString("###,000").PadLeft(7))

            textBox1.ScrollToCaret();// scrolls the textBox to last line

            // draw a small green circle in the center
            CvInvoke.cvCircle(imgOriginal, // draw on the original image
                              new Point((int)circle.Center.X, (int)circle.Center.Y),//center point of circle
                              3, // radios of circle,
                              new MCvScalar (0, 255, 0), // draw in green color
                              -1, //indicates to fill the circle
                              LINE_TYPE.CV_AA, //smoothes the pixels
                              0);// no shift 

            //draw a red circle around the detected object
          //  if (imgOriginal.Data != null) 
                imgOriginal.Draw(circle,    //current circle
                            new Bgr(Color.Red), // draw pure red
                            3);
        } // end of for each

       // if(imgOriginal.Data != null)
            imageBox1.Image = imgOriginal;
    }
于 2013-03-29T12:54:44.197 回答