InAirSignature.exe 中发生类型 'System.IO.FileNotFoundException' 的未处理异常
An unhandled exception of type 'System.IO.FileNotFoundException' occurred in InAirSignature.exe
我正在用 C#Emgu.CV 编写一个程序
我在编译时收到此错误
An unhandled exception of type 'System.IO.FileNotFoundException'
occurred in InAirSignature.exe
Additional information: Could not load file or assembly 'Emgu.CV.UI,
Version=2.9.0.1922, Culture=neutral, PublicKeyToken=7281126722ab4438'
or one of its dependencies. The system cannot find the file specified.
任何想法如何解决？
对于我的编码，需要一个网络摄像头。
下面是我的编码。
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 Emgu.CV;
using Emgu.CV.GPU;
using Emgu.Util;
using Emgu.CV.Structure;
using Emgu.CV.UI;
using Emgu.CV.CvEnum;
using System.Diagnostics;
using System.Windows.Forms.DataVisualization.Charting;
using Emgu.CV.ML.Structure;
using Emgu.CV.ML;
using System.IO;
namespace InAirSignature
{
public partial class Main : Form
{
private Capture capture;
private Rectangle roi;
private Bgr redOrGreen;
private const int NUMBERS_OF_FRAMES = 90,
NUMBERS_OF_SIGN = 6,
NUMBERS_OF_FEATURES = 3, // X Y Averange
FEATURE_X = 0,
FEATURE_Y = 1,
TEST = NUMBERS_OF_SIGN - 1,
NUMBERS_OF_NEURONS = 15;
private const int frameHeight = 640;
private const int frameWidth = 480;
private List<Image<Gray, Byte>> imgGrayBuffer = new List<Image<Gray, Byte>>();
private List<int> pointX = new List<int>();
private List<int> pointY = new List<int>();
private int[][][] pointsOfSign, distanceBetweenPoints;
private double[] finalAverage;
private int[][] pointsOfSignReference;
private int[] distanceBetweenPointsForRefenceAndTest;
private int[] classes = {0,0,0,0,1,1,1,1,1,1};
private int resultMode = 0;
private int recordSign = 0;
private bool record = false;
private int counter = 0;
private Stopwatch stopWatch;
private const string SYSTEM_MESSAGE = "System Message: ",
WELCOME_MESSAGE = "Welcome! ";
Matrix<int> layerSize;
MCvANN_MLP_TrainParams parameters;
ANN_MLP network;
//////////////////////////////////////////////////////////////////////////////////////////////////////////
///////////////////// Form /////////////////////
//////////////////////////////////////////////////////////////////////////////////////////////////////////
#region Form
public Main()
{
InitializeComponent();
#region Initialize graph
chartX.Series.Add("X1");
chartX.Series.Add("X2");
chartX.Series.Add("X3");
chartX.Series.Add("X4");
chartX.Series.Add("X5");
chartX.Series.Add("Xtest");
chartX.Series.Add("Xref");
chartY.Series.Add("Y1");
chartY.Series.Add("Y2");
chartY.Series.Add("Y3");
chartY.Series.Add("Y4");
chartY.Series.Add("Y5");
chartY.Series.Add("Ytest");
chartY.Series.Add("Yref");
chartX.Series["X1"].Color = Color.Blue;
chartX.Series["X2"].Color = Color.Green;
chartX.Series["X3"].Color = Color.Red;
chartX.Series["X4"].Color = Color.Pink;
chartX.Series["X5"].Color = Color.Purple;
chartX.Series["Xtest"].Color = Color.Aqua;
chartX.Series["Xref"].Color = Color.BlueViolet;
chartY.Series["Y1"].Color = Color.Blue;
chartY.Series["Y2"].Color = Color.Green;
chartY.Series["Y3"].Color = Color.Red;
chartY.Series["Y4"].Color = Color.Pink;
chartY.Series["Y5"].Color = Color.Purple;
chartY.Series["Ytest"].Color = Color.Aqua;
chartY.Series["Yref"].Color = Color.BlueViolet;
chartX.Series["X1"].ChartType = SeriesChartType.FastLine;
chartX.Series["X2"].ChartType = SeriesChartType.FastLine;
chartX.Series["X3"].ChartType = SeriesChartType.FastLine;
chartX.Series["X4"].ChartType = SeriesChartType.FastLine;
chartX.Series["X5"].ChartType = SeriesChartType.FastLine;
chartX.Series["Xtest"].ChartType = SeriesChartType.FastLine;
chartX.Series["Xref"].ChartType = SeriesChartType.FastLine;
chartY.Series["Y1"].ChartType = SeriesChartType.FastLine;
chartY.Series["Y2"].ChartType = SeriesChartType.FastLine;
chartY.Series["Y3"].ChartType = SeriesChartType.FastLine;
chartY.Series["Y4"].ChartType = SeriesChartType.FastLine;
chartY.Series["Y5"].ChartType = SeriesChartType.FastLine;
chartY.Series["Ytest"].ChartType = SeriesChartType.FastLine;
chartY.Series["Yref"].ChartType = SeriesChartType.FastLine;
#endregion
#region Initialize Neural Network
layerSize = new Matrix<int>(new int[] { NUMBERS_OF_FEATURES, NUMBERS_OF_NEURONS, 1 });
parameters = new MCvANN_MLP_TrainParams();
parameters.term_crit = new MCvTermCriteria(10, 1.0e-8);
parameters.train_method = Emgu.CV.ML.MlEnum.ANN_MLP_TRAIN_METHOD.BACKPROP;
parameters.bp_dw_scale = 0.1;
parameters.bp_moment_scale = 0.1;
network = new ANN_MLP(layerSize, Emgu.CV.ML.MlEnum.ANN_MLP_ACTIVATION_FUNCTION.SIGMOID_SYM, 1.0, 1.0);
#endregion
#region Initialize camera
capture = new Capture();
capture.SetCaptureProperty(Emgu.CV.CvEnum.CAP_PROP.CV_CAP_PROP_FRAME_HEIGHT, frameHeight);
capture.SetCaptureProperty(Emgu.CV.CvEnum.CAP_PROP.CV_CAP_PROP_FRAME_WIDTH, frameWidth);
#endregion
initializeData();
roi = new Rectangle(338, 2, 300, 300);
redOrGreen = new Bgr(Color.Red);
lblSM.Text = SYSTEM_MESSAGE + WELCOME_MESSAGE;
updateResult();
updateGraph();
Application.Idle += processFrame;
}
private void Form1_FormClosing(object sender, FormClosingEventArgs e)
{
capture.Stop();
}
#endregion
//////////////////////////////////////////////////////////////////////////////////////////////////////////
///////////////////// Button /////////////////////
//////////////////////////////////////////////////////////////////////////////////////////////////////////
#region Button
private void btnSign_Click(object sender, EventArgs e)
{
stopWatch = Stopwatch.StartNew();
record = true;
redOrGreen = new Bgr(Color.LightGreen);
if (sender == btnSign1)
recordSign = 1;
else if (sender == btnSign2)
recordSign = 2;
else if (sender == btnSign3)
recordSign = 3;
else if (sender == btnSign4)
recordSign = 4;
else if (sender == btnSign5)
recordSign = 5;
else
recordSign = 6;
}
private void btnPredict_Click(object sender, EventArgs e)
{
float predicted = predict(new int[,] { { distanceBetweenPointsForRefenceAndTest[0], distanceBetweenPointsForRefenceAndTest[1], distanceBetweenPointsForRefenceAndTest[2] } }, network);
string result;
if (predicted < Convert.ToDouble(lblThreshold.Text))
result = "Success";
else
result = "Fail";
lblSM.Text = SYSTEM_MESSAGE + "Matching result - " + result + ", value: " + predicted.ToString();
}
private void btnSave_Click(object sender, EventArgs e)
{
try
{
if (tbId.Text == "")
throw new Exception();
saveSignReference();
network.Save(tbId.Text);
lblSM.Text = SYSTEM_MESSAGE + "Saved";
}
catch
{
lblSM.Text = SYSTEM_MESSAGE + "ID blank";
}
}
private void btnLoad_Click(object sender, EventArgs e)
{
try
{
if (tbId.Text == "")
throw new Exception();
loadSignReference();
network.Load(tbId.Text);
lblSM.Text = SYSTEM_MESSAGE + "Loaded";
btnVerify.Enabled = true;
btnSave.Enabled = true;
cbAuto.Enabled = true;
updateGraph();
}
catch
{
lblSM.Text = SYSTEM_MESSAGE + "Invalid ID";
}
}
private void btnTrainNN_Click(object sender, EventArgs e)
{
btnVerify.Enabled = true;
btnSave.Enabled = true;
cbAuto.Enabled = true;
trainNN();
}
private void btnClear_Click(object sender, EventArgs e)
{
initializeData();
updateGraph();
updateResult();
calculateResult();
rbS1.PerformClick();
}
#endregion
//////////////////////////////////////////////////////////////////////////////////////////////////////////
///////////////////// Check Box /////////////////////
//////////////////////////////////////////////////////////////////////////////////////////////////////////
#region Check Box
private void cbBinaryImage_CheckedChanged(object sender, EventArgs e)
{
ibBinaryImage.Visible = !ibBinaryImage.Visible;
}
private void cbResult_CheckedChanged(object sender, EventArgs e)
{
pnlResult.Visible = !pnlResult.Visible;
}
private void cbSign_CheckedChanged(object sender, EventArgs e)
{
if (sender == cbSign1)
{
chartX.Series["X1"].Enabled = !chartX.Series["X1"].Enabled;
chartY.Series["Y1"].Enabled = !chartY.Series["Y1"].Enabled;
}
else if (sender == cbSign2)
{
chartX.Series["X2"].Enabled = !chartX.Series["X2"].Enabled;
chartY.Series["Y2"].Enabled = !chartY.Series["Y2"].Enabled;
}
else if (sender == cbSign3)
{
chartX.Series["X3"].Enabled = !chartX.Series["X3"].Enabled;
chartY.Series["Y3"].Enabled = !chartY.Series["Y3"].Enabled;
}
else if (sender == cbSign4)
{
chartX.Series["X4"].Enabled = !chartX.Series["X4"].Enabled;
chartY.Series["Y4"].Enabled = !chartY.Series["Y4"].Enabled;
}
else if (sender == cbSign5)
{
chartX.Series["X5"].Enabled = !chartX.Series["X5"].Enabled;
chartY.Series["Y5"].Enabled = !chartY.Series["Y5"].Enabled;
}
else if (sender == cbSignRefer)
{
chartX.Series["Xref"].Enabled = !chartX.Series["Xref"].Enabled;
chartY.Series["Yref"].Enabled = !chartY.Series["Yref"].Enabled;
}
else
{
chartX.Series["Xtest"].Enabled = !chartX.Series["Xtest"].Enabled;
chartY.Series["Ytest"].Enabled = !chartY.Series["Ytest"].Enabled;
}
}
#endregion
//////////////////////////////////////////////////////////////////////////////////////////////////////////
///////////////////// Radio Button /////////////////////
//////////////////////////////////////////////////////////////////////////////////////////////////////////
#region Radio Button
private void rbS_CheckedChanged(object sender, EventArgs e)
{
if (sender == rbS1)
resultMode = 0;
else if (sender == rbS2)
resultMode = 1;
else if (sender == rbS3)
resultMode = 2;
else if (sender == rbS4)
resultMode = 3;
else
resultMode = 4;
updateResult();
}
#endregion
//////////////////////////////////////////////////////////////////////////////////////////////////////////
///////////////////// Text Box /////////////////////
//////////////////////////////////////////////////////////////////////////////////////////////////////////
#region Text Box
private void tbId_KeyUp(object sender, KeyEventArgs e)
{
if (e.KeyCode == Keys.Enter)
btnLoad.PerformClick();
}
#endregion
//////////////////////////////////////////////////////////////////////////////////////////////////////////
///////////////////// Function /////////////////////
//////////////////////////////////////////////////////////////////////////////////////////////////////////
private void processFrame(object sender, EventArgs arg)
{
Image<Bgr, Byte> frame = capture.QueryFrame().Flip(FLIP.HORIZONTAL);
Image<Bgr, Byte> frameCrop = frame.Copy(new Rectangle(338, 2, 300, 300)); // copy a crop
Image<Gray, Byte> frameBinary;
if (counter == (NUMBERS_OF_FRAMES - 2))
redOrGreen = new Bgr(Color.Red);
frame.Draw(roi, redOrGreen, 2);
frameBinary = ImageProcessing.getSkin(frameCrop);
if (record == true)
{
imgGrayBuffer.Add(frameBinary);
counter++;
if (counter == NUMBERS_OF_FRAMES)
{
while (imgGrayBuffer.Count() > 0)
{
Point point = ImageProcessing.getHighestPoint(imgGrayBuffer.ElementAt(0));
imgGrayBuffer.RemoveAt(0);
pointX.Add(point.X);
pointY.Add(point.Y);
}
pointsOfSign[recordSign - 1][FEATURE_X] = pointX.ToArray();
pointsOfSign[recordSign - 1][FEATURE_X] = normalizeSize(pointsOfSign[recordSign - 1][FEATURE_X]);
pointsOfSign[recordSign - 1][FEATURE_X] = normalizePosition(pointsOfSign[recordSign - 1][FEATURE_X]);
pointsOfSign[recordSign - 1][FEATURE_Y] = pointY.ToArray();
pointsOfSign[recordSign - 1][FEATURE_Y] = normalizeSize(pointsOfSign[recordSign - 1][FEATURE_Y]);
pointsOfSign[recordSign - 1][FEATURE_Y] = normalizePosition(pointsOfSign[recordSign - 1][FEATURE_Y]);
pointX.Clear();
pointY.Clear();
calculateResult();
updateResult();
updateGraph();
record = false;
stopWatch.Stop();
lblSM.Text = SYSTEM_MESSAGE + "Time taken => " + stopWatch.Elapsed.TotalSeconds.ToString() + " seconds.";
if (cbAuto.Checked == true && recordSign == 6)
btnVerify.PerformClick();
recordSign = 0;
counter = 0;
}
}
ibMain.Image = frame;
if(cbBinaryImage.Checked == true)
ibBinaryImage.Image = frameBinary;
}
private int[] normalizePosition(int[] position)
{
int min = int.MaxValue,
max = int.MinValue;
for (int i = 0; i < position.Count(); i++)
{
if (position[i] > max)
max = position[i];
if (position[i] < min)
min = position[i];
}
int midPoint = (min + max) / 2;
for (int i = 0; i < position.Count(); i++)
position[i] -= midPoint;
return position;
}
private int[] normalizeSize(int[] position)
{
const int targetSize = 300;
int min = int.MaxValue,
max = int.MinValue;
for (int i = 0; i < position.Count(); i++)
{
if (position[i] > max)
max = position[i];
if (position[i] < min)
min = position[i];
}
int height = max - min;
if (height != 0)
height = targetSize / height;
for (int i = 0; i < position.Count(); i++)
position[i] *= height;
return position;
}
private void updateGraph()
{
foreach (var series in chartX.Series)
series.Points.Clear();
foreach (var series in chartY.Series)
series.Points.Clear();
for (int i = 0; i < NUMBERS_OF_FRAMES; i++)
{
chartX.Series["X1"].Points.AddXY
(i, pointsOfSign[0][FEATURE_X][i]);
chartX.Series["X2"].Points.AddXY
(i, pointsOfSign[1][FEATURE_X][i]);
chartX.Series["X3"].Points.AddXY
(i, pointsOfSign[2][FEATURE_X][i]);
chartX.Series["X4"].Points.AddXY
(i, pointsOfSign[3][FEATURE_X][i]);
chartX.Series["X5"].Points.AddXY
(i, pointsOfSign[4][FEATURE_X][i]);
chartX.Series["Xref"].Points.AddXY
(i, pointsOfSignReference[0][i]);
chartX.Series["Xtest"].Points.AddXY
(i, pointsOfSign[5][FEATURE_X][i]);
chartY.Series["Y1"].Points.AddXY
(i, pointsOfSign[0][FEATURE_Y][i]);
chartY.Series["Y2"].Points.AddXY
(i, pointsOfSign[1][FEATURE_Y][i]);
chartY.Series["Y3"].Points.AddXY
(i, pointsOfSign[2][FEATURE_Y][i]);
chartY.Series["Y4"].Points.AddXY
(i, pointsOfSign[3][FEATURE_Y][i]);
chartY.Series["Y5"].Points.AddXY
(i, pointsOfSign[4][FEATURE_Y][i]);
chartY.Series["Yref"].Points.AddXY
(i, pointsOfSignReference[1][i]);
chartY.Series["Ytest"].Points.AddXY
(i, pointsOfSign[5][FEATURE_Y][i]);
}
}
private void calculateResult()
{
int average = 0;
for (int i = 0; i < NUMBERS_OF_SIGN - 1; i++)
for (int j = 0; j < NUMBERS_OF_SIGN - 1; j++)
{
for (int k = 0; k < NUMBERS_OF_FEATURES - 1; k++)
{
distanceBetweenPoints[i][j][k] = DTW.Distance(pointsOfSign[i][k], pointsOfSign[j][k]);
average += distanceBetweenPoints[i][j][k];
}
distanceBetweenPoints[i][j][NUMBERS_OF_FEATURES - 1] = average / 2;
average = 0;
}
for (int i = 0; i < NUMBERS_OF_SIGN -1; i++)
{
finalAverage[i] = 0.0;
for (int j = 0; j < NUMBERS_OF_SIGN; j++)
finalAverage[i] += distanceBetweenPoints[i][j][NUMBERS_OF_FEATURES - 1];
finalAverage[i] /= NUMBERS_OF_SIGN - 2;
}
average = 0;
for (int k = 0; k < NUMBERS_OF_FEATURES - 1; k++)
{
distanceBetweenPointsForRefenceAndTest[k] = DTW.Distance(pointsOfSignReference[k], pointsOfSign[TEST][k]);
average += distanceBetweenPointsForRefenceAndTest[k];
}
distanceBetweenPointsForRefenceAndTest[NUMBERS_OF_FEATURES - 1] = average / 2;
}
private void updateResult()
{
tbResult1X.Text = distanceBetweenPoints[resultMode][0][0].ToString();
tbResult1Y.Text = distanceBetweenPoints[resultMode][0][1].ToString();
tbResult1A.Text = distanceBetweenPoints[resultMode][0][2].ToString();
tbResult2X.Text = distanceBetweenPoints[resultMode][1][0].ToString();
tbResult2Y.Text = distanceBetweenPoints[resultMode][1][1].ToString();
tbResult2A.Text = distanceBetweenPoints[resultMode][1][2].ToString();
tbResult3X.Text = distanceBetweenPoints[resultMode][2][0].ToString();
tbResult3Y.Text = distanceBetweenPoints[resultMode][2][1].ToString();
tbResult3A.Text = distanceBetweenPoints[resultMode][2][2].ToString();
tbResult4X.Text = distanceBetweenPoints[resultMode][3][0].ToString();
tbResult4Y.Text = distanceBetweenPoints[resultMode][3][1].ToString();
tbResult4A.Text = distanceBetweenPoints[resultMode][3][2].ToString();
tbResult5X.Text = distanceBetweenPoints[resultMode][4][0].ToString();
tbResult5Y.Text = distanceBetweenPoints[resultMode][4][1].ToString();
tbResult5A.Text = distanceBetweenPoints[resultMode][4][2].ToString();
tbTotalA.Text = finalAverage[resultMode].ToString();
tbRatX.Text = distanceBetweenPointsForRefenceAndTest[0].ToString();
tbRatY.Text = distanceBetweenPointsForRefenceAndTest[1].ToString();
tbRatA.Text = distanceBetweenPointsForRefenceAndTest[2].ToString();
}
private float predict(int[,] testingSetInt, ANN_MLP network)
{
Matrix<float> testingSet = new Matrix<float>(Utility.arrayIntToFloat(testingSetInt)),
prediction = new Matrix<float>(1, 1);
network.Predict(testingSet, prediction);
return prediction.Data[0, 0];
}
private ANN_MLP trainNN(int[] trainingClassesInt, int[,] trainingSetInt)
{
int numbers_of_training_set = 4,
numbers_of_data = trainingSetInt.GetUpperBound(0) + 1;
Matrix<float> trainingSet = new Matrix<float>(Utility.arrayIntToFloat(trainingSetInt)),
trainingClasses = new Matrix<float>(Utility.arrayIntToFloat(trainingClassesInt));
Matrix<float> trainingSetPositive = trainingSet.GetRows(0, numbers_of_training_set, 1);
Matrix<float> trainingSetNegative = trainingSet.GetRows(numbers_of_training_set, numbers_of_data, 1);
Matrix<float> trainClassesPositive = trainingClasses.GetRows(0, numbers_of_training_set, 1);
Matrix<float> trainClassesNegative = trainingClasses.GetRows(numbers_of_training_set, numbers_of_data, 1);
ANN_MLP network = new ANN_MLP(layerSize, Emgu.CV.ML.MlEnum.ANN_MLP_ACTIVATION_FUNCTION.SIGMOID_SYM, 1.0, 1.0);
network.Train(trainingSet, trainingClasses, null, (Matrix<int>)null, parameters, Emgu.CV.ML.MlEnum.ANN_MLP_TRAINING_FLAG.DEFAULT);
return network;
}
private void trainNN()
{
// Getting the most closest signature
int index = 0;
for (int i = 1; i < NUMBERS_OF_SIGN - 1; i++)
if (finalAverage[index] > finalAverage[i])
index = i;
lblSM.Text = SYSTEM_MESSAGE + "Sign @ no." + (index + 1).ToString() + " is the reference now!";
for (int i = 0; i < NUMBERS_OF_FEATURES - 1; i++)
pointsOfSignReference[i] = pointsOfSign[index][i];
int total_training_data = 10; // 4 positive, 6 negative
int total_features_for_training = 3; // dtw of x, dtw of y, width of sign
int[,] trainingSet = new int[total_training_data, total_features_for_training];
// Get the distance (x and y) between closest signature and the rest of signature into training set
// Feature 0 and 1 => dtw of x, dtw of y
for (int i = 0, current = 0; i < NUMBERS_OF_SIGN - 1; i++)
{
if (i != index)
{
for (int j = 0; j < NUMBERS_OF_FEATURES; j++)
trainingSet[current, j] = distanceBetweenPoints[i][index][j];
current++;
}
}
// Generate negative data
int[] max_value = new int[NUMBERS_OF_FEATURES];
int[] min_value = new int[NUMBERS_OF_FEATURES];
for (int i = 0; i < NUMBERS_OF_SIGN - 1; i++)
{
for (int j = 0; j < NUMBERS_OF_FEATURES && i != index; j++)
{
if (max_value[j] < distanceBetweenPoints[i][index][j])
max_value[j] = distanceBetweenPoints[i][index][j];
if (min_value[j] > distanceBetweenPoints[i][index][j])
min_value[j] = distanceBetweenPoints[i][index][j];
}
}
Random random = new Random();
在您的项目中，右键单击“引用”，然后查找 Emgu.CV.UI 程序集并 select 它。
我正在用 C#Emgu.CV 编写一个程序
我在编译时收到此错误
An unhandled exception of type 'System.IO.FileNotFoundException' occurred in InAirSignature.exe
Additional information: Could not load file or assembly 'Emgu.CV.UI, Version=2.9.0.1922, Culture=neutral, PublicKeyToken=7281126722ab4438' or one of its dependencies. The system cannot find the file specified.
任何想法如何解决？ 对于我的编码，需要一个网络摄像头。 下面是我的编码。
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 Emgu.CV;
using Emgu.CV.GPU;
using Emgu.Util;
using Emgu.CV.Structure;
using Emgu.CV.UI;
using Emgu.CV.CvEnum;
using System.Diagnostics;
using System.Windows.Forms.DataVisualization.Charting;
using Emgu.CV.ML.Structure;
using Emgu.CV.ML;
using System.IO;
namespace InAirSignature
{
public partial class Main : Form
{
private Capture capture;
private Rectangle roi;
private Bgr redOrGreen;
private const int NUMBERS_OF_FRAMES = 90,
NUMBERS_OF_SIGN = 6,
NUMBERS_OF_FEATURES = 3, // X Y Averange
FEATURE_X = 0,
FEATURE_Y = 1,
TEST = NUMBERS_OF_SIGN - 1,
NUMBERS_OF_NEURONS = 15;
private const int frameHeight = 640;
private const int frameWidth = 480;
private List<Image<Gray, Byte>> imgGrayBuffer = new List<Image<Gray, Byte>>();
private List<int> pointX = new List<int>();
private List<int> pointY = new List<int>();
private int[][][] pointsOfSign, distanceBetweenPoints;
private double[] finalAverage;
private int[][] pointsOfSignReference;
private int[] distanceBetweenPointsForRefenceAndTest;
private int[] classes = {0,0,0,0,1,1,1,1,1,1};
private int resultMode = 0;
private int recordSign = 0;
private bool record = false;
private int counter = 0;
private Stopwatch stopWatch;
private const string SYSTEM_MESSAGE = "System Message: ",
WELCOME_MESSAGE = "Welcome! ";
Matrix<int> layerSize;
MCvANN_MLP_TrainParams parameters;
ANN_MLP network;
//////////////////////////////////////////////////////////////////////////////////////////////////////////
///////////////////// Form /////////////////////
//////////////////////////////////////////////////////////////////////////////////////////////////////////
#region Form
public Main()
{
InitializeComponent();
#region Initialize graph
chartX.Series.Add("X1");
chartX.Series.Add("X2");
chartX.Series.Add("X3");
chartX.Series.Add("X4");
chartX.Series.Add("X5");
chartX.Series.Add("Xtest");
chartX.Series.Add("Xref");
chartY.Series.Add("Y1");
chartY.Series.Add("Y2");
chartY.Series.Add("Y3");
chartY.Series.Add("Y4");
chartY.Series.Add("Y5");
chartY.Series.Add("Ytest");
chartY.Series.Add("Yref");
chartX.Series["X1"].Color = Color.Blue;
chartX.Series["X2"].Color = Color.Green;
chartX.Series["X3"].Color = Color.Red;
chartX.Series["X4"].Color = Color.Pink;
chartX.Series["X5"].Color = Color.Purple;
chartX.Series["Xtest"].Color = Color.Aqua;
chartX.Series["Xref"].Color = Color.BlueViolet;
chartY.Series["Y1"].Color = Color.Blue;
chartY.Series["Y2"].Color = Color.Green;
chartY.Series["Y3"].Color = Color.Red;
chartY.Series["Y4"].Color = Color.Pink;
chartY.Series["Y5"].Color = Color.Purple;
chartY.Series["Ytest"].Color = Color.Aqua;
chartY.Series["Yref"].Color = Color.BlueViolet;
chartX.Series["X1"].ChartType = SeriesChartType.FastLine;
chartX.Series["X2"].ChartType = SeriesChartType.FastLine;
chartX.Series["X3"].ChartType = SeriesChartType.FastLine;
chartX.Series["X4"].ChartType = SeriesChartType.FastLine;
chartX.Series["X5"].ChartType = SeriesChartType.FastLine;
chartX.Series["Xtest"].ChartType = SeriesChartType.FastLine;
chartX.Series["Xref"].ChartType = SeriesChartType.FastLine;
chartY.Series["Y1"].ChartType = SeriesChartType.FastLine;
chartY.Series["Y2"].ChartType = SeriesChartType.FastLine;
chartY.Series["Y3"].ChartType = SeriesChartType.FastLine;
chartY.Series["Y4"].ChartType = SeriesChartType.FastLine;
chartY.Series["Y5"].ChartType = SeriesChartType.FastLine;
chartY.Series["Ytest"].ChartType = SeriesChartType.FastLine;
chartY.Series["Yref"].ChartType = SeriesChartType.FastLine;
#endregion
#region Initialize Neural Network
layerSize = new Matrix<int>(new int[] { NUMBERS_OF_FEATURES, NUMBERS_OF_NEURONS, 1 });
parameters = new MCvANN_MLP_TrainParams();
parameters.term_crit = new MCvTermCriteria(10, 1.0e-8);
parameters.train_method = Emgu.CV.ML.MlEnum.ANN_MLP_TRAIN_METHOD.BACKPROP;
parameters.bp_dw_scale = 0.1;
parameters.bp_moment_scale = 0.1;
network = new ANN_MLP(layerSize, Emgu.CV.ML.MlEnum.ANN_MLP_ACTIVATION_FUNCTION.SIGMOID_SYM, 1.0, 1.0);
#endregion
#region Initialize camera
capture = new Capture();
capture.SetCaptureProperty(Emgu.CV.CvEnum.CAP_PROP.CV_CAP_PROP_FRAME_HEIGHT, frameHeight);
capture.SetCaptureProperty(Emgu.CV.CvEnum.CAP_PROP.CV_CAP_PROP_FRAME_WIDTH, frameWidth);
#endregion
initializeData();
roi = new Rectangle(338, 2, 300, 300);
redOrGreen = new Bgr(Color.Red);
lblSM.Text = SYSTEM_MESSAGE + WELCOME_MESSAGE;
updateResult();
updateGraph();
Application.Idle += processFrame;
}
private void Form1_FormClosing(object sender, FormClosingEventArgs e)
{
capture.Stop();
}
#endregion
//////////////////////////////////////////////////////////////////////////////////////////////////////////
///////////////////// Button /////////////////////
//////////////////////////////////////////////////////////////////////////////////////////////////////////
#region Button
private void btnSign_Click(object sender, EventArgs e)
{
stopWatch = Stopwatch.StartNew();
record = true;
redOrGreen = new Bgr(Color.LightGreen);
if (sender == btnSign1)
recordSign = 1;
else if (sender == btnSign2)
recordSign = 2;
else if (sender == btnSign3)
recordSign = 3;
else if (sender == btnSign4)
recordSign = 4;
else if (sender == btnSign5)
recordSign = 5;
else
recordSign = 6;
}
private void btnPredict_Click(object sender, EventArgs e)
{
float predicted = predict(new int[,] { { distanceBetweenPointsForRefenceAndTest[0], distanceBetweenPointsForRefenceAndTest[1], distanceBetweenPointsForRefenceAndTest[2] } }, network);
string result;
if (predicted < Convert.ToDouble(lblThreshold.Text))
result = "Success";
else
result = "Fail";
lblSM.Text = SYSTEM_MESSAGE + "Matching result - " + result + ", value: " + predicted.ToString();
}
private void btnSave_Click(object sender, EventArgs e)
{
try
{
if (tbId.Text == "")
throw new Exception();
saveSignReference();
network.Save(tbId.Text);
lblSM.Text = SYSTEM_MESSAGE + "Saved";
}
catch
{
lblSM.Text = SYSTEM_MESSAGE + "ID blank";
}
}
private void btnLoad_Click(object sender, EventArgs e)
{
try
{
if (tbId.Text == "")
throw new Exception();
loadSignReference();
network.Load(tbId.Text);
lblSM.Text = SYSTEM_MESSAGE + "Loaded";
btnVerify.Enabled = true;
btnSave.Enabled = true;
cbAuto.Enabled = true;
updateGraph();
}
catch
{
lblSM.Text = SYSTEM_MESSAGE + "Invalid ID";
}
}
private void btnTrainNN_Click(object sender, EventArgs e)
{
btnVerify.Enabled = true;
btnSave.Enabled = true;
cbAuto.Enabled = true;
trainNN();
}
private void btnClear_Click(object sender, EventArgs e)
{
initializeData();
updateGraph();
updateResult();
calculateResult();
rbS1.PerformClick();
}
#endregion
//////////////////////////////////////////////////////////////////////////////////////////////////////////
///////////////////// Check Box /////////////////////
//////////////////////////////////////////////////////////////////////////////////////////////////////////
#region Check Box
private void cbBinaryImage_CheckedChanged(object sender, EventArgs e)
{
ibBinaryImage.Visible = !ibBinaryImage.Visible;
}
private void cbResult_CheckedChanged(object sender, EventArgs e)
{
pnlResult.Visible = !pnlResult.Visible;
}
private void cbSign_CheckedChanged(object sender, EventArgs e)
{
if (sender == cbSign1)
{
chartX.Series["X1"].Enabled = !chartX.Series["X1"].Enabled;
chartY.Series["Y1"].Enabled = !chartY.Series["Y1"].Enabled;
}
else if (sender == cbSign2)
{
chartX.Series["X2"].Enabled = !chartX.Series["X2"].Enabled;
chartY.Series["Y2"].Enabled = !chartY.Series["Y2"].Enabled;
}
else if (sender == cbSign3)
{
chartX.Series["X3"].Enabled = !chartX.Series["X3"].Enabled;
chartY.Series["Y3"].Enabled = !chartY.Series["Y3"].Enabled;
}
else if (sender == cbSign4)
{
chartX.Series["X4"].Enabled = !chartX.Series["X4"].Enabled;
chartY.Series["Y4"].Enabled = !chartY.Series["Y4"].Enabled;
}
else if (sender == cbSign5)
{
chartX.Series["X5"].Enabled = !chartX.Series["X5"].Enabled;
chartY.Series["Y5"].Enabled = !chartY.Series["Y5"].Enabled;
}
else if (sender == cbSignRefer)
{
chartX.Series["Xref"].Enabled = !chartX.Series["Xref"].Enabled;
chartY.Series["Yref"].Enabled = !chartY.Series["Yref"].Enabled;
}
else
{
chartX.Series["Xtest"].Enabled = !chartX.Series["Xtest"].Enabled;
chartY.Series["Ytest"].Enabled = !chartY.Series["Ytest"].Enabled;
}
}
#endregion
//////////////////////////////////////////////////////////////////////////////////////////////////////////
///////////////////// Radio Button /////////////////////
//////////////////////////////////////////////////////////////////////////////////////////////////////////
#region Radio Button
private void rbS_CheckedChanged(object sender, EventArgs e)
{
if (sender == rbS1)
resultMode = 0;
else if (sender == rbS2)
resultMode = 1;
else if (sender == rbS3)
resultMode = 2;
else if (sender == rbS4)
resultMode = 3;
else
resultMode = 4;
updateResult();
}
#endregion
//////////////////////////////////////////////////////////////////////////////////////////////////////////
///////////////////// Text Box /////////////////////
//////////////////////////////////////////////////////////////////////////////////////////////////////////
#region Text Box
private void tbId_KeyUp(object sender, KeyEventArgs e)
{
if (e.KeyCode == Keys.Enter)
btnLoad.PerformClick();
}
#endregion
//////////////////////////////////////////////////////////////////////////////////////////////////////////
///////////////////// Function /////////////////////
//////////////////////////////////////////////////////////////////////////////////////////////////////////
private void processFrame(object sender, EventArgs arg)
{
Image<Bgr, Byte> frame = capture.QueryFrame().Flip(FLIP.HORIZONTAL);
Image<Bgr, Byte> frameCrop = frame.Copy(new Rectangle(338, 2, 300, 300)); // copy a crop
Image<Gray, Byte> frameBinary;
if (counter == (NUMBERS_OF_FRAMES - 2))
redOrGreen = new Bgr(Color.Red);
frame.Draw(roi, redOrGreen, 2);
frameBinary = ImageProcessing.getSkin(frameCrop);
if (record == true)
{
imgGrayBuffer.Add(frameBinary);
counter++;
if (counter == NUMBERS_OF_FRAMES)
{
while (imgGrayBuffer.Count() > 0)
{
Point point = ImageProcessing.getHighestPoint(imgGrayBuffer.ElementAt(0));
imgGrayBuffer.RemoveAt(0);
pointX.Add(point.X);
pointY.Add(point.Y);
}
pointsOfSign[recordSign - 1][FEATURE_X] = pointX.ToArray();
pointsOfSign[recordSign - 1][FEATURE_X] = normalizeSize(pointsOfSign[recordSign - 1][FEATURE_X]);
pointsOfSign[recordSign - 1][FEATURE_X] = normalizePosition(pointsOfSign[recordSign - 1][FEATURE_X]);
pointsOfSign[recordSign - 1][FEATURE_Y] = pointY.ToArray();
pointsOfSign[recordSign - 1][FEATURE_Y] = normalizeSize(pointsOfSign[recordSign - 1][FEATURE_Y]);
pointsOfSign[recordSign - 1][FEATURE_Y] = normalizePosition(pointsOfSign[recordSign - 1][FEATURE_Y]);
pointX.Clear();
pointY.Clear();
calculateResult();
updateResult();
updateGraph();
record = false;
stopWatch.Stop();
lblSM.Text = SYSTEM_MESSAGE + "Time taken => " + stopWatch.Elapsed.TotalSeconds.ToString() + " seconds.";
if (cbAuto.Checked == true && recordSign == 6)
btnVerify.PerformClick();
recordSign = 0;
counter = 0;
}
}
ibMain.Image = frame;
if(cbBinaryImage.Checked == true)
ibBinaryImage.Image = frameBinary;
}
private int[] normalizePosition(int[] position)
{
int min = int.MaxValue,
max = int.MinValue;
for (int i = 0; i < position.Count(); i++)
{
if (position[i] > max)
max = position[i];
if (position[i] < min)
min = position[i];
}
int midPoint = (min + max) / 2;
for (int i = 0; i < position.Count(); i++)
position[i] -= midPoint;
return position;
}
private int[] normalizeSize(int[] position)
{
const int targetSize = 300;
int min = int.MaxValue,
max = int.MinValue;
for (int i = 0; i < position.Count(); i++)
{
if (position[i] > max)
max = position[i];
if (position[i] < min)
min = position[i];
}
int height = max - min;
if (height != 0)
height = targetSize / height;
for (int i = 0; i < position.Count(); i++)
position[i] *= height;
return position;
}
private void updateGraph()
{
foreach (var series in chartX.Series)
series.Points.Clear();
foreach (var series in chartY.Series)
series.Points.Clear();
for (int i = 0; i < NUMBERS_OF_FRAMES; i++)
{
chartX.Series["X1"].Points.AddXY
(i, pointsOfSign[0][FEATURE_X][i]);
chartX.Series["X2"].Points.AddXY
(i, pointsOfSign[1][FEATURE_X][i]);
chartX.Series["X3"].Points.AddXY
(i, pointsOfSign[2][FEATURE_X][i]);
chartX.Series["X4"].Points.AddXY
(i, pointsOfSign[3][FEATURE_X][i]);
chartX.Series["X5"].Points.AddXY
(i, pointsOfSign[4][FEATURE_X][i]);
chartX.Series["Xref"].Points.AddXY
(i, pointsOfSignReference[0][i]);
chartX.Series["Xtest"].Points.AddXY
(i, pointsOfSign[5][FEATURE_X][i]);
chartY.Series["Y1"].Points.AddXY
(i, pointsOfSign[0][FEATURE_Y][i]);
chartY.Series["Y2"].Points.AddXY
(i, pointsOfSign[1][FEATURE_Y][i]);
chartY.Series["Y3"].Points.AddXY
(i, pointsOfSign[2][FEATURE_Y][i]);
chartY.Series["Y4"].Points.AddXY
(i, pointsOfSign[3][FEATURE_Y][i]);
chartY.Series["Y5"].Points.AddXY
(i, pointsOfSign[4][FEATURE_Y][i]);
chartY.Series["Yref"].Points.AddXY
(i, pointsOfSignReference[1][i]);
chartY.Series["Ytest"].Points.AddXY
(i, pointsOfSign[5][FEATURE_Y][i]);
}
}
private void calculateResult()
{
int average = 0;
for (int i = 0; i < NUMBERS_OF_SIGN - 1; i++)
for (int j = 0; j < NUMBERS_OF_SIGN - 1; j++)
{
for (int k = 0; k < NUMBERS_OF_FEATURES - 1; k++)
{
distanceBetweenPoints[i][j][k] = DTW.Distance(pointsOfSign[i][k], pointsOfSign[j][k]);
average += distanceBetweenPoints[i][j][k];
}
distanceBetweenPoints[i][j][NUMBERS_OF_FEATURES - 1] = average / 2;
average = 0;
}
for (int i = 0; i < NUMBERS_OF_SIGN -1; i++)
{
finalAverage[i] = 0.0;
for (int j = 0; j < NUMBERS_OF_SIGN; j++)
finalAverage[i] += distanceBetweenPoints[i][j][NUMBERS_OF_FEATURES - 1];
finalAverage[i] /= NUMBERS_OF_SIGN - 2;
}
average = 0;
for (int k = 0; k < NUMBERS_OF_FEATURES - 1; k++)
{
distanceBetweenPointsForRefenceAndTest[k] = DTW.Distance(pointsOfSignReference[k], pointsOfSign[TEST][k]);
average += distanceBetweenPointsForRefenceAndTest[k];
}
distanceBetweenPointsForRefenceAndTest[NUMBERS_OF_FEATURES - 1] = average / 2;
}
private void updateResult()
{
tbResult1X.Text = distanceBetweenPoints[resultMode][0][0].ToString();
tbResult1Y.Text = distanceBetweenPoints[resultMode][0][1].ToString();
tbResult1A.Text = distanceBetweenPoints[resultMode][0][2].ToString();
tbResult2X.Text = distanceBetweenPoints[resultMode][1][0].ToString();
tbResult2Y.Text = distanceBetweenPoints[resultMode][1][1].ToString();
tbResult2A.Text = distanceBetweenPoints[resultMode][1][2].ToString();
tbResult3X.Text = distanceBetweenPoints[resultMode][2][0].ToString();
tbResult3Y.Text = distanceBetweenPoints[resultMode][2][1].ToString();
tbResult3A.Text = distanceBetweenPoints[resultMode][2][2].ToString();
tbResult4X.Text = distanceBetweenPoints[resultMode][3][0].ToString();
tbResult4Y.Text = distanceBetweenPoints[resultMode][3][1].ToString();
tbResult4A.Text = distanceBetweenPoints[resultMode][3][2].ToString();
tbResult5X.Text = distanceBetweenPoints[resultMode][4][0].ToString();
tbResult5Y.Text = distanceBetweenPoints[resultMode][4][1].ToString();
tbResult5A.Text = distanceBetweenPoints[resultMode][4][2].ToString();
tbTotalA.Text = finalAverage[resultMode].ToString();
tbRatX.Text = distanceBetweenPointsForRefenceAndTest[0].ToString();
tbRatY.Text = distanceBetweenPointsForRefenceAndTest[1].ToString();
tbRatA.Text = distanceBetweenPointsForRefenceAndTest[2].ToString();
}
private float predict(int[,] testingSetInt, ANN_MLP network)
{
Matrix<float> testingSet = new Matrix<float>(Utility.arrayIntToFloat(testingSetInt)),
prediction = new Matrix<float>(1, 1);
network.Predict(testingSet, prediction);
return prediction.Data[0, 0];
}
private ANN_MLP trainNN(int[] trainingClassesInt, int[,] trainingSetInt)
{
int numbers_of_training_set = 4,
numbers_of_data = trainingSetInt.GetUpperBound(0) + 1;
Matrix<float> trainingSet = new Matrix<float>(Utility.arrayIntToFloat(trainingSetInt)),
trainingClasses = new Matrix<float>(Utility.arrayIntToFloat(trainingClassesInt));
Matrix<float> trainingSetPositive = trainingSet.GetRows(0, numbers_of_training_set, 1);
Matrix<float> trainingSetNegative = trainingSet.GetRows(numbers_of_training_set, numbers_of_data, 1);
Matrix<float> trainClassesPositive = trainingClasses.GetRows(0, numbers_of_training_set, 1);
Matrix<float> trainClassesNegative = trainingClasses.GetRows(numbers_of_training_set, numbers_of_data, 1);
ANN_MLP network = new ANN_MLP(layerSize, Emgu.CV.ML.MlEnum.ANN_MLP_ACTIVATION_FUNCTION.SIGMOID_SYM, 1.0, 1.0);
network.Train(trainingSet, trainingClasses, null, (Matrix<int>)null, parameters, Emgu.CV.ML.MlEnum.ANN_MLP_TRAINING_FLAG.DEFAULT);
return network;
}
private void trainNN()
{
// Getting the most closest signature
int index = 0;
for (int i = 1; i < NUMBERS_OF_SIGN - 1; i++)
if (finalAverage[index] > finalAverage[i])
index = i;
lblSM.Text = SYSTEM_MESSAGE + "Sign @ no." + (index + 1).ToString() + " is the reference now!";
for (int i = 0; i < NUMBERS_OF_FEATURES - 1; i++)
pointsOfSignReference[i] = pointsOfSign[index][i];
int total_training_data = 10; // 4 positive, 6 negative
int total_features_for_training = 3; // dtw of x, dtw of y, width of sign
int[,] trainingSet = new int[total_training_data, total_features_for_training];
// Get the distance (x and y) between closest signature and the rest of signature into training set
// Feature 0 and 1 => dtw of x, dtw of y
for (int i = 0, current = 0; i < NUMBERS_OF_SIGN - 1; i++)
{
if (i != index)
{
for (int j = 0; j < NUMBERS_OF_FEATURES; j++)
trainingSet[current, j] = distanceBetweenPoints[i][index][j];
current++;
}
}
// Generate negative data
int[] max_value = new int[NUMBERS_OF_FEATURES];
int[] min_value = new int[NUMBERS_OF_FEATURES];
for (int i = 0; i < NUMBERS_OF_SIGN - 1; i++)
{
for (int j = 0; j < NUMBERS_OF_FEATURES && i != index; j++)
{
if (max_value[j] < distanceBetweenPoints[i][index][j])
max_value[j] = distanceBetweenPoints[i][index][j];
if (min_value[j] > distanceBetweenPoints[i][index][j])
min_value[j] = distanceBetweenPoints[i][index][j];
}
}
Random random = new Random();
在您的项目中，右键单击“引用”，然后查找 Emgu.CV.UI 程序集并 select 它。