// Use this for initialization
void Start()
{
cascade = CvHaarClassifierCascade.FromFile(@"./Assets/haarcascade_frontalface_alt.xml");
capture = Cv.CreateCameraCapture(0);
Cv.SetCaptureProperty(capture, CaptureProperty.FrameWidth, CAPTURE_WIDTH);
Cv.SetCaptureProperty(capture, CaptureProperty.FrameHeight, CAPTURE_HEIGHT);
IplImage frame = Cv.QueryFrame(capture);
Cv.NamedWindow("FaceDetect");
CvSVM svm = new CvSVM ();
CvTermCriteria criteria = new CvTermCriteria (CriteriaType.Epsilon, 1000, double.Epsilon);
CvSVMParams param = new CvSVMParams (CvSVM.C_SVC, CvSVM.RBF, 10.0, 8.0, 1.0, 10.0, 0.5, 0.1, null, criteria);
}
public SVM()
{
// CvSVM
// SVMを利用して2次元ベクトルの3クラス分類問題を解く
const int S = 1000;
const int SIZE = 400;
CvRNG rng = new CvRNG((ulong)DateTime.Now.Ticks);
// (1)画像領域の確保と初期化
using (IplImage img = new IplImage(SIZE, SIZE, BitDepth.U8, 3))
{
img.Zero();
// (2)学習データの生成
CvPoint[] pts = new CvPoint[S];
int[] res = new int[S];
for (int i = 0; i < S; i++)
{
pts[i].X = (int)(rng.RandInt() % SIZE);
pts[i].Y = (int)(rng.RandInt() % SIZE);
if (pts[i].Y > 50 * Math.Cos(pts[i].X * Cv.PI / 100) + 200)
{
img.Line(new CvPoint(pts[i].X - 2, pts[i].Y - 2), new CvPoint(pts[i].X + 2, pts[i].Y + 2), new CvColor(255, 0, 0));
img.Line(new CvPoint(pts[i].X + 2, pts[i].Y - 2), new CvPoint(pts[i].X - 2, pts[i].Y + 2), new CvColor(255, 0, 0));
res[i] = 1;
}
else
{
if (pts[i].X > 200)
{
img.Line(new CvPoint(pts[i].X - 2, pts[i].Y - 2), new CvPoint(pts[i].X + 2, pts[i].Y + 2), new CvColor(0, 255, 0));
img.Line(new CvPoint(pts[i].X + 2, pts[i].Y - 2), new CvPoint(pts[i].X - 2, pts[i].Y + 2), new CvColor(0, 255, 0));
res[i] = 2;
}
else
{
img.Line(new CvPoint(pts[i].X - 2, pts[i].Y - 2), new CvPoint(pts[i].X + 2, pts[i].Y + 2), new CvColor(0, 0, 255));
img.Line(new CvPoint(pts[i].X + 2, pts[i].Y - 2), new CvPoint(pts[i].X - 2, pts[i].Y + 2), new CvColor(0, 0, 255));
res[i] = 3;
}
}
}
// (3)学習データの表示
Cv.NamedWindow("SVM", WindowMode.AutoSize);
Cv.ShowImage("SVM", img);
Cv.WaitKey(0);
// (4)学習パラメータの生成
float[] data = new float[S * 2];
for (int i = 0; i < S; i++)
{
data[i * 2] = ((float)pts[i].X) / SIZE;
data[i * 2 + 1] = ((float)pts[i].Y) / SIZE;
}
// (5)SVMの学習
using (CvSVM svm = new CvSVM())
{
CvMat data_mat = new CvMat(S, 2, MatrixType.F32C1, data);
CvMat res_mat = new CvMat(S, 1, MatrixType.S32C1, res);
CvTermCriteria criteria = new CvTermCriteria(1000, float.Epsilon);
CvSVMParams param = new CvSVMParams(SVMType.CSvc, SVMKernelType.Rbf, 10.0, 8.0, 1.0, 10.0, 0.5, 0.1, null, criteria);
svm.Train(data_mat, res_mat, null, null, param);
// (6)学習結果の描画
for (int i = 0; i < SIZE; i++)
{
for (int j = 0; j < SIZE; j++)
{
float[] a = { (float)j / SIZE, (float)i / SIZE };
CvMat m = new CvMat(1, 2, MatrixType.F32C1, a);
float ret = svm.Predict(m);
CvColor color = new CvColor();
switch ((int)ret)
{
case 1:
color = new CvColor(100, 0, 0); break;
case 2:
color = new CvColor(0, 100, 0); break;
case 3:
color = new CvColor(0, 0, 100); break;
}
img[i, j] = color;
}
}
// (7)トレーニングデータの再描画
for (int i = 0; i < S; i++)
{
CvColor color = new CvColor();
switch (res[i])
{
case 1:
color = new CvColor(255, 0, 0); break;
case 2:
color = new CvColor(0, 255, 0); break;
case 3:
color = new CvColor(0, 0, 255); break;
}
img.Line(new CvPoint(pts[i].X - 2, pts[i].Y - 2), new CvPoint(pts[i].X + 2, pts[i].Y + 2), color);
img.Line(new CvPoint(pts[i].X + 2, pts[i].Y - 2), new CvPoint(pts[i].X - 2, pts[i].Y + 2), color);
}
// (8)サポートベクターの描画
int sv_num = svm.GetSupportVectorCount();
for (int i = 0; i < sv_num; i++)
{
var support = svm.GetSupportVector(i);
img.Circle(new CvPoint((int)(support[0] * SIZE), (int)(support[1] * SIZE)), 5, new CvColor(200, 200, 200));
}
// (9)画像の表示
Cv.NamedWindow("SVM", WindowMode.AutoSize);
Cv.ShowImage("SVM", img);
Cv.WaitKey(0);
Cv.DestroyWindow("SVM");
}
}
}
/// <summary>
/// 初期化
/// </summary>
/// <param name="train_data"></param>
/// <param name="responses"></param>
/// <param name="var_idx"></param>
/// <param name="sample_idx"></param>
/// <param name="params"></param>
#else
/// <summary>
/// Training constructor
/// </summary>
/// <param name="trainData"></param>
/// <param name="responses"></param>
/// <param name="varIdx"></param>
/// <param name="sampleIdx"></param>
/// <param name="params"></param>
#endif
public CvSVM(CvMat trainData, CvMat responses, CvMat varIdx, CvMat sampleIdx, CvSVMParams @params)
{
if (trainData == null)
throw new ArgumentNullException("trainData");
if (responses == null)
throw new ArgumentNullException("responses");
if(@params == null)
@params = new CvSVMParams();
IntPtr varIdxPtr = (varIdx == null) ? IntPtr.Zero : varIdx.CvPtr;
IntPtr sampleIdxPtr = (sampleIdx == null) ? IntPtr.Zero : sampleIdx.CvPtr;
ptr = MLInvoke.CvSVM_construct_training(
trainData.CvPtr,
responses.CvPtr,
varIdxPtr,
sampleIdxPtr,
@params.NativeStruct
);
NotifyMemoryPressure(SizeOf);
}
/// <summary>
/// SVMを最適なパラメータで学習する
/// </summary>
/// <param name="train_data"></param>
/// <param name="responses"></param>
/// <param name="var_idx"></param>
/// <param name="sample_idx"></param>
/// <param name="params"></param>
/// <param name="k_fold">交差検定（Cross-validation）パラメータ．学習集合は，k_foldの部分集合に分割され，一つの部分集合がモデルの学習に用いられ，その他の部分集合はテスト集合となる．つまり，SVM アルゴリズムは，k_fold回実行される.</param>
/// <param name="C_grid"></param>
/// <param name="gamma_grid"></param>
/// <param name="p_grid"></param>
/// <param name="nu_grid"></param>
/// <param name="coef_grid"></param>
/// <param name="degree_grid"></param>
/// <returns></returns>
#else
/// <summary>
/// Trains SVM with optimal parameters
/// </summary>
/// <param name="trainData"></param>
/// <param name="responses"></param>
/// <param name="varIdx"></param>
/// <param name="sampleIdx"></param>
/// <param name="params"></param>
/// <param name="kFold">Cross-validation parameter. The training set is divided into k_fold subsets, one subset being used to train the model, the others forming the test set. So, the SVM algorithm is executed k_fold times. </param>
/// <param name="cGrid"></param>
/// <param name="gammaGrid"></param>
/// <param name="pGrid"></param>
/// <param name="nuGrid"></param>
/// <param name="coefGrid"></param>
/// <param name="degreeGrid"></param>
/// <returns></returns>
#endif
public virtual bool TrainAuto(CvMat trainData, CvMat responses, CvMat varIdx, CvMat sampleIdx, CvSVMParams @params, int kFold,
CvParamGrid cGrid, CvParamGrid gammaGrid, CvParamGrid pGrid, CvParamGrid nuGrid, CvParamGrid coefGrid, CvParamGrid degreeGrid)
{
if (trainData == null)
throw new ArgumentNullException("trainData");
if (responses == null)
throw new ArgumentNullException("responses");
if(@params == null)
@params = new CvSVMParams();
IntPtr varIdxPtr = (varIdx == null) ? IntPtr.Zero : varIdx.CvPtr;
IntPtr sampleIdxPtr = (sampleIdx == null) ? IntPtr.Zero : sampleIdx.CvPtr;
return MLInvoke.CvSVM_train_auto(
ptr,
trainData.CvPtr,
responses.CvPtr,
varIdxPtr,
sampleIdxPtr,
@params.NativeStruct,
kFold,
cGrid,
gammaGrid,
pGrid,
nuGrid,
coefGrid,
degreeGrid
);
}
/// <summary>
/// SVMを最適なパラメータで学習する
/// </summary>
/// <param name="train_data"></param>
/// <param name="responses"></param>
/// <param name="var_idx"></param>
/// <param name="sample_idx"></param>
/// <param name="params"></param>
/// <param name="k_fold">交差検定（Cross-validation）パラメータ．学習集合は，k_foldの部分集合に分割され，一つの部分集合がモデルの学習に用いられ，その他の部分集合はテスト集合となる．つまり，SVM アルゴリズムは，k_fold回実行される.</param>
/// <returns></returns>
#else
/// <summary>
/// Trains SVM with optimal parameters
/// </summary>
/// <param name="trainData"></param>
/// <param name="responses"></param>
/// <param name="varIdx"></param>
/// <param name="sampleIdx"></param>
/// <param name="params"></param>
/// <param name="kFold">Cross-validation parameter. The training set is divided into k_fold subsets, one subset being used to train the model, the others forming the test set. So, the SVM algorithm is executed k_fold times. </param>
/// <returns></returns>
#endif
public virtual bool TrainAuto(CvMat trainData, CvMat responses, CvMat varIdx, CvMat sampleIdx, CvSVMParams @params, int kFold)
{
return TrainAuto(trainData, responses, varIdx, sampleIdx, @params, kFold,
GetDefaultGrid(SVMParamType.C), GetDefaultGrid(SVMParamType.Gamma),GetDefaultGrid(SVMParamType.P),
GetDefaultGrid(SVMParamType.Nu),GetDefaultGrid(SVMParamType.Coef),GetDefaultGrid(SVMParamType.Degree));
}
/// <summary>
/// SVMを最適なパラメータで学習する
/// </summary>
/// <param name="train_data"></param>
/// <param name="responses"></param>
/// <param name="var_idx"></param>
/// <param name="sample_idx"></param>
/// <param name="params"></param>
/// <returns></returns>
#else
/// <summary>
/// Trains SVM with optimal parameters
/// </summary>
/// <param name="trainData"></param>
/// <param name="responses"></param>
/// <param name="varIdx"></param>
/// <param name="sampleIdx"></param>
/// <param name="params"></param>
/// <returns></returns>
#endif
public virtual bool TrainAuto(CvMat trainData, CvMat responses, CvMat varIdx, CvMat sampleIdx, CvSVMParams @params)
{
return TrainAuto(trainData, responses, varIdx, sampleIdx, @params, 10);
}
/// <summary>
/// SVMを学習する
/// </summary>
/// <param name="train_data"></param>
/// <param name="responses"></param>
/// <param name="var_idx"></param>
/// <param name="sample_idx"></param>
/// <param name="params"></param>
#else
/// <summary>
/// Trains SVM
/// </summary>
/// <param name="trainData"></param>
/// <param name="responses"></param>
/// <param name="varIdx"></param>
/// <param name="sampleIdx"></param>
/// <param name="params"></param>
#endif
public virtual bool Train(CvMat trainData, CvMat responses, CvMat varIdx, CvMat sampleIdx, CvSVMParams @params)
{
if (trainData == null)
throw new ArgumentNullException("trainData");
if (responses == null)
throw new ArgumentNullException("responses");
if(@params == null)
@params = new CvSVMParams();
IntPtr varIdxPtr = (varIdx == null) ? IntPtr.Zero : varIdx.CvPtr;
IntPtr sampleIdxPtr = (sampleIdx == null) ? IntPtr.Zero : sampleIdx.CvPtr;
return MLInvoke.CvSVM_train(
ptr,
trainData.CvPtr,
responses.CvPtr,
varIdxPtr,
sampleIdxPtr,
@params.NativeStruct
);
}
