| public Train ( IImage images, int labels ) : void | ||
| images | IImage | The images used in the training |
| labels | int | The labels of the images |
| 리턴 | void | |
public Form1()
{
InitializeComponent();
recognizer = new LBPHFaceRecognizer(1, 8, 8, 9, 65);
classifier = new CascadeClassifier(haarcascade);
GPU_classifier = new GpuCascadeClassifier(haarcascade_cuda);
font = new MCvFont(Emgu.CV.CvEnum.FONT.CV_FONT_HERSHEY_TRIPLEX, 0.5, 0.5);
if (File.Exists(@"traningdata.xml"))
{
recognizer.Load(@"traningdata.xml");
}
else
{
foreach (var file in Directory.GetFiles(Application.StartupPath + @"\Traning Faces\"))
{
try { temp = new Image<Gray, Byte>(file); }
catch { continue; }
temp._EqualizeHist();
var detectedFaces = classifier.DetectMultiScale(temp, 1.1, 15, new Size(24, 24), Size.Empty);
if (detectedFaces.Length == 0)
{
continue;
}
temp.ROI = detectedFaces[0];
temp = temp.Copy();
temp = temp.Resize(100, 100, Emgu.CV.CvEnum.INTER.CV_INTER_CUBIC);
imagesList.Add(temp);
imagesLabels.Add(Path.GetFileNameWithoutExtension(file));
}
for (int i = 0; i < imagesList.Count; i++)
{
imagesLabels_indices.Add(i);
}
try { recognizer.Train(imagesList.ToArray(), imagesLabels_indices.ToArray()); }
catch (Exception ex)
{
MessageBox.Show(ex.Message);
Environment.Exit(0);
}
}
}
private bool LoadTrainingData()
{
mydb = new DBConn();
allname = mydb.getLabelList();
trainingImages = mydb.getTrainedImageList();
int[] temp = Enumerable.Range(0,(allname.Count)).ToArray();
if (mydb.getImageCount() > 0)
{
if (trainingImages.Length != 0)
{
//set round and ...
//termCrit = new MCvTermCriteria(mydb.getImageCount(), 0.001);
//Eigen face recognizer
recognizer = new FisherFaceRecognizer(0,3200);//4000
recognizer.Train(trainingImages, temp);
return true;
}
else
{
return false;
}
}
else
{
return false;
}
}
/// <summary>
/// Loads the traing data given a (string) folder location
/// </summary>
/// <param name="Folder_location"></param>
/// <returns></returns>
private bool LoadTrainingData(string Folder_location)
{
if (File.Exists(Folder_location + "\\TrainedLabels.xml"))
{
try
{
//message_bar.Text = "";
Names_List.Clear();
Names_List_ID.Clear();
trainingImages.Clear();
FileStream filestream = File.OpenRead(Folder_location + "\\TrainedLabels.xml");
long filelength = filestream.Length;
byte[] xmlBytes = new byte[filelength];
filestream.Read(xmlBytes, 0, (int)filelength);
filestream.Close();
MemoryStream xmlStream = new MemoryStream(xmlBytes);
using (XmlReader xmlreader = XmlTextReader.Create(xmlStream))
{
while (xmlreader.Read())
{
if (xmlreader.IsStartElement())
{
switch (xmlreader.Name)
{
case "NAME":
if (xmlreader.Read())
{
Names_List_ID.Add(Names_List.Count); //0, 1, 2, 3....
Names_List.Add(xmlreader.Value.Trim());
NumLabels += 1;
}
break;
case "FILE":
if (xmlreader.Read())
{
//PROBLEM HERE IF TRAININGG MOVED
trainingImages.Add(new Image<Gray, byte>(Application.StartupPath + "\\TrainedFaces\\" + xmlreader.Value.Trim()));
}
break;
}
}
}
}
ContTrain = NumLabels;
if (trainingImages.ToArray().Length != 0)
{
//Eigen face recognizer
//Parameters:
// num_components – The number of components (read: Eigenfaces) kept for this Prinicpal
// Component Analysis. As a hint: There’s no rule how many components (read: Eigenfaces)
// should be kept for good reconstruction capabilities. It is based on your input data,
// so experiment with the number. Keeping 80 components should almost always be sufficient.
//
// threshold – The threshold applied in the prediciton. This still has issues as it work inversly to LBH and Fisher Methods.
// if you use 0.0 recognizer.Predict will always return -1 or unknown if you use 5000 for example unknow won't be reconised.
// As in previous versions I ignore the built in threhold methods and allow a match to be found i.e. double.PositiveInfinity
// and then use the eigen distance threshold that is return to elliminate unknowns.
//
//NOTE: The following causes the confusion, sinc two rules are used.
//--------------------------------------------------------------------------------------------------------------------------------------
//Eigen Uses
// 0 - X = unknown
// > X = Recognised
//
//Fisher and LBPH Use
// 0 - X = Recognised
// > X = Unknown
//
// Where X = Threshold value
switch (Recognizer_Type)
{
case ("EMGU.CV.LBPHFaceRecognizer"):
recognizer = new LBPHFaceRecognizer(1, 8, 8, 8, 100);//50
break;
case ("EMGU.CV.FisherFaceRecognizer"):
recognizer = new FisherFaceRecognizer(0, 3500);//4000
break;
case ("EMGU.CV.EigenFaceRecognizer"):
default:
recognizer = new EigenFaceRecognizer(80, double.PositiveInfinity);
break;
}
recognizer.Train(trainingImages.ToArray(), Names_List_ID.ToArray());
// Recognizer_Type = recognizer.GetType();
// string v = recognizer.ToString(); //EMGU.CV.FisherFaceRecognizer || EMGU.CV.EigenFaceRecognizer || EMGU.CV.LBPHFaceRecognizer
return true;
}
else return false;
}
catch (Exception ex)
{
Error = ex.ToString();
return false;
}
}
else return false;
}
private bool loadTrainingData(string folderLocation)
{
if (File.Exists(folderLocation + "/TrainedLabels.xml"))
{
try
{
// Purge all existing data
namesList.Clear();
trainingImages.Clear();
// Read the XML file into memory
FileStream fileStream = File.OpenRead(folderLocation + "/TrainedLabels.xml");
long fileLength = fileStream.Length;
byte[] xmlBytes = new byte[fileLength];
fileStream.Read(xmlBytes, 0, (int)fileLength);
fileStream.Close();
MemoryStream xmlStream = new MemoryStream(xmlBytes);
using (XmlReader xmlReader = XmlTextReader.Create(xmlStream))
{
while (xmlReader.Read())
{
if (xmlReader.IsStartElement())
{
switch (xmlReader.Name)
{
case "NAME":
if (xmlReader.Read())
{
// Add the name to the data set
namesList.Add(xmlReader.Value.Trim());
// Store an identifier. This is merely the position in namesList
namesListId.Add((namesList.Count));
// Increment the counter for entries in the data set
numLabels += 1;
}
break;
case "FILE":
if (xmlReader.Read())
{
// Add the image to the data set
trainingImages.Add(new Image<Gray, Byte>(System.AppDomain.CurrentDomain.BaseDirectory + "/TrainedFaces/" + xmlReader.Value.Trim()));
}
break;
}
}
}
}
if (trainingImages.ToArray().Length != 0)
{
recognizer = new LBPHFaceRecognizer(1, 8, 8, 9, 100);
recognizer.Train(trainingImages.ToArray(), namesListId.ToArray());
return true;
}
else
{
return false;
}
}
catch (Exception ex)
{
error = ex.ToString();
return false;
}
}
else
{
return false;
}
}