/// <summary>
/// 決定木を学習する
/// </summary>
/// <param name="trainData"></param>
/// <param name="subsampleIdx"></param>
/// <param name="ensemble"></param>
/// <returns></returns>
#else
/// <summary>
/// Trains decision tree
/// </summary>
/// <param name="trainData"></param>
/// <param name="subsampleIdx"></param>
/// <param name="ensemble"></param>
/// <returns></returns>
#endif
public virtual bool Train(CvDTreeTrainData trainData, CvMat subsampleIdx, CvBoost ensemble)
{
if (trainData == null)
{
throw new ArgumentNullException("trainData");
}
if (subsampleIdx == null)
{
throw new ArgumentNullException("subsampleIdx");
}
return(NativeMethods.ml_CvBoostTree_train(
ptr, trainData.CvPtr, subsampleIdx.CvPtr, Cv2.ToPtr(ensemble)) != 0);
}
/// <summary>
/// サンプル集合からガウス混合パラメータを推定する
/// </summary>
/// <param name="samples"></param>
/// <param name="logLikelihoods"></param>
/// <param name="labels"></param>
/// <param name="probs"></param>
#else
/// <summary>
/// Estimates Gaussian mixture parameters from the sample set
/// </summary>
/// <param name="samples"></param>
/// <param name="logLikelihoods"></param>
/// <param name="labels"></param>
/// <param name="probs"></param>
#endif
public virtual bool Train(
InputArray samples,
OutputArray logLikelihoods = null,
OutputArray labels = null,
OutputArray probs = null)
{
if (disposed)
{
throw new ObjectDisposedException("EM");
}
if (samples == null)
{
throw new ArgumentNullException("samples");
}
samples.ThrowIfDisposed();
if (logLikelihoods != null)
{
logLikelihoods.ThrowIfNotReady();
}
if (labels != null)
{
labels.ThrowIfNotReady();
}
if (probs != null)
{
probs.ThrowIfNotReady();
}
int ret = NativeMethods.ml_EM_train(
ptr,
samples.CvPtr,
Cv2.ToPtr(logLikelihoods),
Cv2.ToPtr(labels),
Cv2.ToPtr(probs));
if (logLikelihoods != null)
{
logLikelihoods.Fix();
}
if (labels != null)
{
labels.Fix();
}
if (probs != null)
{
probs.Fix();
}
return(ret != 0);
}
/// <summary>
///
/// </summary>
/// <param name="defaultMat"></param>
/// <returns></returns>
public SparseMat ReadSparseMat(SparseMat defaultMat = null)
{
var value = new SparseMat();
try
{
NativeMethods.core_FileNode_read_SparseMat(ptr, value.CvPtr, Cv2.ToPtr(defaultMat));
}
catch
{
value.Dispose();
throw;
}
return(value);
}
/// <summary>
/// 高速なマルチスケール Hesian 検出器を用いて keypoint を検出します．
/// </summary>
/// <param name="img"></param>
/// <param name="mask"></param>
/// <returns></returns>
#else
/// <summary>
/// detects keypoints using fast multi-scale Hessian detector
/// </summary>
/// <param name="img"></param>
/// <param name="mask"></param>
/// <returns></returns>
#endif
public KeyPoint[] Run(InputArray img, Mat mask)
{
ThrowIfDisposed();
if (img == null)
{
throw new ArgumentNullException("img");
}
img.ThrowIfDisposed();
using (VectorOfKeyPoint keypointsVec = new VectorOfKeyPoint())
{
NativeMethods.nonfree_SURF_run1(ptr, img.CvPtr, Cv2.ToPtr(mask), keypointsVec.CvPtr);
return(keypointsVec.ToArray());
}
}
/// <summary>
/// Compute the BRISK features on an image
/// </summary>
/// <param name="image"></param>
/// <param name="mask"></param>
/// <returns></returns>
public KeyPoint[] Run(InputArray image, InputArray mask = null)
{
ThrowIfDisposed();
if (image == null)
{
throw new ArgumentNullException("image");
}
image.ThrowIfDisposed();
using (VectorOfKeyPoint keyPointsVec = new VectorOfKeyPoint())
{
NativeMethods.features2d_BRISK_run1(ptr, image.CvPtr, Cv2.ToPtr(mask), keyPointsVec.CvPtr);
return(keyPointsVec.ToArray());
}
}
/// <summary>
/// 二つの学習サンプル間の近さを取り出す
/// </summary>
/// <param name="sample1"></param>
/// <param name="sample2"></param>
/// <param name="missing1"></param>
/// <param name="missing2"></param>
/// <returns></returns>
#else
/// <summary>
/// Retrieves proximity measure between two training samples
/// </summary>
/// <param name="sample1"></param>
/// <param name="sample2"></param>
/// <param name="missing1"></param>
/// <param name="missing2"></param>
/// <returns></returns>
#endif
public virtual float GetProximity(
CvMat sample1, CvMat sample2,
CvMat missing1 = null, CvMat missing2 = null)
{
if (sample1 == null)
{
throw new ArgumentNullException("sample1");
}
if (sample2 == null)
{
throw new ArgumentNullException("sample2");
}
return(NativeMethods.ml_CvRTrees_get_proximity(
ptr, sample1.CvPtr, sample2.CvPtr,
Cv2.ToPtr(missing1), Cv2.ToPtr(missing2)));
}
/// <summary>
/// MSERのすべての輪郭情報を抽出する
/// </summary>
/// <param name="image"></param>
/// <param name="mask"></param>
/// <returns></returns>
#else
/// <summary>
/// Extracts the contours of Maximally Stable Extremal Regions
/// </summary>
/// <param name="image"></param>
/// <param name="mask"></param>
/// <returns></returns>
#endif
public Point[][] Run(Mat image, Mat mask)
{
ThrowIfDisposed();
if (image == null)
{
throw new ArgumentNullException("image");
}
image.ThrowIfDisposed();
IntPtr msers;
NativeMethods.features2d_MSER_detect(ptr, image.CvPtr, out msers, Cv2.ToPtr(mask));
using (VectorOfVectorPoint msersVec = new VectorOfVectorPoint(msers))
{
return(msersVec.ToArray());
}
}
/// <summary>
/// Find one best match for each query descriptor (if mask is empty).
/// </summary>
/// <param name="queryDescriptors"></param>
/// <param name="trainDescriptors"></param>
/// <param name="mask"></param>
/// <returns></returns>
public DMatch[] Match(Mat queryDescriptors, Mat trainDescriptors, Mat mask = null)
{
ThrowIfDisposed();
if (queryDescriptors == null)
{
throw new ArgumentNullException(nameof(queryDescriptors));
}
if (trainDescriptors == null)
{
throw new ArgumentNullException(nameof(trainDescriptors));
}
using (var matchesVec = new VectorOfDMatch())
{
NativeMethods.features2d_DescriptorMatcher_match1(
ptr, queryDescriptors.CvPtr, trainDescriptors.CvPtr,
matchesVec.CvPtr, Cv2.ToPtr(mask));
return(matchesVec.ToArray());
}
}
/// <summary>
///
/// </summary>
/// <param name="trainData"></param>
/// <param name="tflag"></param>
/// <param name="responses"></param>
/// <param name="varIdx"></param>
/// <param name="sampleIdx"></param>
/// <param name="varType"></param>
/// <param name="missingMask"></param>
/// <param name="param"></param>
/// <param name="shared"></param>
/// <param name="addLabels"></param>
/// <param name="updateData"></param>
/// <returns></returns>
#else
/// <summary>
///
/// </summary>
/// <param name="trainData"></param>
/// <param name="tflag"></param>
/// <param name="responses"></param>
/// <param name="varIdx"></param>
/// <param name="sampleIdx"></param>
/// <param name="varType"></param>
/// <param name="missingMask"></param>
/// <param name="param"></param>
/// <param name="shared"></param>
/// <param name="addLabels"></param>
/// <param name="updateData"></param>
/// <returns></returns>
#endif
public void SetData(
CvMat trainData,
DTreeDataLayout tflag,
CvMat responses,
CvMat varIdx = null,
CvMat sampleIdx = null,
CvMat varType = null,
CvMat missingMask = null,
CvDTreeParams param = null,
bool shared = false,
bool addLabels = false,
bool updateData = false)
{
if (trainData == null)
{
throw new ArgumentNullException(nameof(trainData));
}
if (responses == null)
{
throw new ArgumentNullException(nameof(responses));
}
if (param == null)
{
param = new CvDTreeParams();
}
NativeMethods.ml_CvDTreeTrainData_set_data(
ptr,
trainData.CvPtr,
(int)tflag,
responses.CvPtr,
Cv2.ToPtr(varIdx),
Cv2.ToPtr(sampleIdx),
Cv2.ToPtr(varType),
Cv2.ToPtr(missingMask),
param.CvPtr,
shared ? 1 : 0,
addLabels ? 1 : 0,
updateData ? 1 : 0
);
}
/// <summary>
/// keypoint を検出し，その SURF ディスクリプタを計算します．[useProvidedKeypoints = true]
/// </summary>
/// <param name="img"></param>
/// <param name="mask"></param>
/// <param name="keypoints"></param>
/// <param name="descriptors"></param>
/// <param name="useProvidedKeypoints"></param>
#else
/// <summary>
/// detects keypoints and computes the SURF descriptors for them. [useProvidedKeypoints = true]
/// </summary>
/// <param name="img"></param>
/// <param name="mask"></param>
/// <param name="keypoints"></param>
/// <param name="descriptors"></param>
/// <param name="useProvidedKeypoints"></param>
#endif
public void Run(InputArray img, InputArray mask, out KeyPoint[] keypoints, out float[] descriptors,
bool useProvidedKeypoints = false)
{
ThrowIfDisposed();
if (img == null)
{
throw new ArgumentNullException("img");
}
img.ThrowIfDisposed();
using (VectorOfKeyPoint keypointsVec = new VectorOfKeyPoint())
using (VectorOfFloat descriptorsVec = new VectorOfFloat())
{
NativeMethods.nonfree_SURF_run2_vector(ptr, img.CvPtr, Cv2.ToPtr(mask), keypointsVec.CvPtr,
descriptorsVec.CvPtr, useProvidedKeypoints ? 1 : 0);
keypoints = keypointsVec.ToArray();
descriptors = descriptorsVec.ToArray();
}
}
/// <summary>
/// Find k best matches for each query descriptor (in increasing order of distances).
/// compactResult is used when mask is not empty. If compactResult is false matches
/// vector will have the same size as queryDescriptors rows. If compactResult is true
/// matches vector will not contain matches for fully masked out query descriptors.
/// </summary>
/// <param name="queryDescriptors"></param>
/// <param name="trainDescriptors"></param>
/// <param name="k"></param>
/// <param name="mask"></param>
/// <param name="compactResult"></param>
/// <returns></returns>
public DMatch[][] KnnMatch(Mat queryDescriptors, Mat trainDescriptors,
int k, Mat mask = null, bool compactResult = false)
{
ThrowIfDisposed();
if (queryDescriptors == null)
{
throw new ArgumentNullException("queryDescriptors");
}
if (trainDescriptors == null)
{
throw new ArgumentNullException("trainDescriptors");
}
using (VectorOfVectorDMatch matchesVec = new VectorOfVectorDMatch())
{
NativeMethods.features2d_DescriptorMatcher_knnMatch(
ptr, queryDescriptors.CvPtr, trainDescriptors.CvPtr,
matchesVec.CvPtr, k, Cv2.ToPtr(mask), compactResult ? 1 : 0);
return(matchesVec.ToArray());
}
}
/// <summary>
/// 入力ベクトルに対する決定木の葉ノードを返す
/// </summary>
/// <param name="sample"></param>
/// <param name="missingDataMask"></param>
/// <param name="preprocessedInput">falseは通常の入力を意味する．true の場合，このメソッドは離散入力変数の全ての値があらかじめ， 0..<num_of_categories_i>-1 の範囲に正規化されていることを仮定する（決定木は内部的にはこのような正規化された表現を用いている）．これは決定木集合の高速な予測に役立つ．連続変数の入力変数に対しては，このフラグは利用されない．</param>
/// <returns></returns>
#else
/// <summary>
/// Returns the leaf node of decision tree corresponding to the input vector
/// </summary>
/// <param name="sample"></param>
/// <param name="missingDataMask"></param>
/// <param name="preprocessedInput"></param>
/// <returns></returns>
#endif
public virtual CvDTreeNode Predict(
Mat sample,
Mat missingDataMask = null,
bool preprocessedInput = false)
{
if (sample == null)
{
throw new ArgumentNullException(nameof(sample));
}
sample.ThrowIfDisposed();
IntPtr result = NativeMethods.ml_CvDTree_predict_Mat(
ptr, sample.CvPtr, Cv2.ToPtr(missingDataMask), preprocessedInput ? 1 : 0);
if (result == IntPtr.Zero)
{
return(null);
}
return(new CvDTreeNode(result));
}
/// <summary>
/// Find best matches for each query descriptor which have distance less than
/// maxDistance (in increasing order of distances).
/// </summary>
/// <param name="queryDescriptors"></param>
/// <param name="trainDescriptors"></param>
/// <param name="maxDistance"></param>
/// <param name="mask"></param>
/// <param name="compactResult"></param>
/// <returns></returns>
public DMatch[][] RadiusMatch(Mat queryDescriptors, Mat trainDescriptors,
float maxDistance, Mat mask = null, bool compactResult = false)
{
ThrowIfDisposed();
if (queryDescriptors == null)
{
throw new ArgumentNullException(nameof(queryDescriptors));
}
if (trainDescriptors == null)
{
throw new ArgumentNullException(nameof(trainDescriptors));
}
using (var matchesVec = new VectorOfVectorDMatch())
{
NativeMethods.features2d_DescriptorMatcher_radiusMatch1(
ptr, queryDescriptors.CvPtr, trainDescriptors.CvPtr,
matchesVec.CvPtr, maxDistance, Cv2.ToPtr(mask), compactResult ? 1 : 0);
return(matchesVec.ToArray());
}
}
/// <summary>
/// 学習データを与えて初期化
/// </summary>
/// <param name="trainData">既知のサンプル (m*n)</param>
/// <param name="responses">既知のサンプルのクラス (m*1)</param>
/// <param name="varIdx"></param>
/// <param name="sampleIdx"></param>
/// <returns></returns>
#else
/// <summary>
/// Bayes classifier for normally distributed data
/// </summary>
/// <param name="trainData">Known samples (m*n)</param>
/// <param name="responses">Classes for known samples (m*1)</param>
/// <param name="varIdx"></param>
/// <param name="sampleIdx"></param>
/// <returns></returns>
#endif
public CvNormalBayesClassifier(
Mat trainData,
Mat responses,
Mat varIdx = null,
Mat sampleIdx = null)
{
if (trainData == null)
{
throw new ArgumentNullException("trainData");
}
if (responses == null)
{
throw new ArgumentNullException("responses");
}
trainData.ThrowIfDisposed();
responses.ThrowIfDisposed();
ptr = NativeMethods.ml_CvNormalBayesClassifier_new2_Mat(
trainData.CvPtr, responses.CvPtr, Cv2.ToPtr(varIdx), Cv2.ToPtr(sampleIdx));
}
/// <summary>
/// 学習データを与えて初期化
/// </summary>
/// <param name="trainData"></param>
/// <param name="tflag"></param>
/// <param name="responses"></param>
/// <param name="varIdx"></param>
/// <param name="sampleIdx"></param>
/// <param name="varType"></param>
/// <param name="missingMask"></param>
/// <param name="param"></param>
/// <param name="shared"></param>
/// <param name="addLabels"></param>
/// <returns></returns>
#else
/// <summary>
/// Training constructor
/// </summary>
/// <param name="trainData"></param>
/// <param name="tflag"></param>
/// <param name="responses"></param>
/// <param name="varIdx"></param>
/// <param name="sampleIdx"></param>
/// <param name="varType"></param>
/// <param name="missingMask"></param>
/// <param name="param"></param>
/// <param name="shared"></param>
/// <param name="addLabels"></param>
/// <returns></returns>
#endif
public CvDTreeTrainData(
CvMat trainData,
DTreeDataLayout tflag,
CvMat responses,
CvMat varIdx = null,
CvMat sampleIdx = null,
CvMat varType = null,
CvMat missingMask = null,
CvDTreeParams param = null,
bool shared = false,
bool addLabels = false)
{
if (trainData == null)
{
throw new ArgumentNullException("trainData");
}
if (responses == null)
{
throw new ArgumentNullException("responses");
}
trainData.ThrowIfDisposed();
responses.ThrowIfDisposed();
if (param == null)
{
param = new CvDTreeParams();
}
ptr = NativeMethods.ml_CvDTreeTrainData_new2(
trainData.CvPtr,
(int)tflag,
responses.CvPtr,
Cv2.ToPtr(varIdx),
Cv2.ToPtr(sampleIdx),
Cv2.ToPtr(varType),
Cv2.ToPtr(missingMask),
param.CvPtr,
shared ? 1 : 0,
addLabels ? 1 : 0
);
}
/// <summary>
/// ブーストされた分類器の学習
/// </summary>
/// <param name="trainData"></param>
/// <param name="tflag"></param>
/// <param name="responses"></param>
/// <param name="varIdx"></param>
/// <param name="sampleIdx"></param>
/// <param name="varType"></param>
/// <param name="missingMask"></param>
/// <param name="param"></param>
/// <param name="update"></param>
/// <returns></returns>
#else
/// <summary>
/// Trains boosted tree classifier
/// </summary>
/// <param name="trainData"></param>
/// <param name="tflag"></param>
/// <param name="responses"></param>
/// <param name="varIdx"></param>
/// <param name="sampleIdx"></param>
/// <param name="varType"></param>
/// <param name="missingMask"></param>
/// <param name="param"></param>
/// <param name="update"></param>
/// <returns></returns>
#endif
public virtual bool Train(
Mat trainData,
DTreeDataLayout tflag,
Mat responses,
Mat varIdx = null,
Mat sampleIdx = null,
Mat varType = null,
Mat missingMask = null,
CvBoostParams param = null,
bool update = false)
{
if (trainData == null)
{
throw new ArgumentNullException("trainData");
}
if (responses == null)
{
throw new ArgumentNullException("responses");
}
trainData.ThrowIfDisposed();
responses.ThrowIfDisposed();
if (param == null)
{
param = new CvBoostParams();
}
int ret = NativeMethods.ml_CvBoost_train_Mat(
ptr,
trainData.CvPtr,
(int)tflag,
responses.CvPtr,
Cv2.ToPtr(varIdx),
Cv2.ToPtr(sampleIdx),
Cv2.ToPtr(varType),
Cv2.ToPtr(missingMask),
param.CvPtr,
update ? 1 : 0);
return(ret != 0);
}
/// <summary>
/// 入力サンプルに対する応答を予測する
/// </summary>
/// <param name="sample">入力サンプル</param>
/// <param name="missing">データ欠損マスク（オプション）．データ欠損を扱うためには，弱い決定木が代理分岐を含まなければならない．</param>
/// <param name="slice">予測に用いられる弱い決定木シーケンスの連続的部分集合（スライス）．デフォルトでは，全ての弱い分類器が用いられる．</param>
/// <param name="rawMode">falseは通常の入力を意味する．true の場合，このメソッドは離散入力変数の全ての値があらかじめ， 0..<num_of_categories_i>-1 の範囲に正規化されていることを仮定する（決定木は内部的にはこのような正規化された表現を用いている）．これは決定木集合の高速な予測に役立つ．連続変数の入力変数に対しては，このフラグは利用されない．</param>
/// <param name="returnSum"></param>
/// <returns>重み付き投票に基づく出力クラスラベル</returns>
#else
/// <summary>
/// Predicts response for the input sample
/// </summary>
/// <param name="sample">The input sample. </param>
/// <param name="missing">The optional mask of missing measurements. To handle missing measurements, the weak classifiers must include surrogate splits. </param>
/// <param name="slice">The continuous subset of the sequence of weak classifiers to be used for prediction. By default, all the weak classifiers are used. </param>
/// <param name="rawMode">The last parameter is normally set to false that implies a regular input. If it is true, the method assumes that all the values of the discrete input variables have been already normalized to 0..<num_of_categoriesi>-1 ranges. (as the decision tree uses such normalized representation internally). It is useful for faster prediction with tree ensembles. For ordered input variables the flag is not used. </param>
/// <param name="returnSum"></param>
/// <returns>the output class label based on the weighted voting. </returns>
#endif
public float Predict(
Mat sample,
Mat missing = null,
Range?slice = null,
bool rawMode = false,
bool returnSum = false)
{
if (sample == null)
{
throw new ArgumentNullException("sample");
}
CvSlice slice0 = slice.GetValueOrDefault(CvSlice.WholeSeq);
return(NativeMethods.ml_CvBoost_predict_Mat(
ptr,
sample.CvPtr,
Cv2.ToPtr(missing),
slice0,
rawMode ? 1 : 0,
returnSum ? 1 : 0));
}
/// <summary>
/// keypoint を検出し，その SURF ディスクリプタを計算します．[useProvidedKeypoints = true]
/// </summary>
/// <param name="img"></param>
/// <param name="mask"></param>
/// <param name="keypoints"></param>
/// <param name="descriptors"></param>
/// <param name="useProvidedKeypoints"></param>
#else
/// <summary>
/// detects keypoints and computes the SURF descriptors for them. [useProvidedKeypoints = true]
/// </summary>
/// <param name="img"></param>
/// <param name="mask"></param>
/// <param name="keypoints"></param>
/// <param name="descriptors"></param>
/// <param name="useProvidedKeypoints"></param>
#endif
public void Run(InputArray img, InputArray mask, out KeyPoint[] keypoints, OutputArray descriptors,
bool useProvidedKeypoints = false)
{
ThrowIfDisposed();
if (img == null)
{
throw new ArgumentNullException(nameof(img));
}
if (descriptors == null)
{
throw new ArgumentNullException(nameof(descriptors));
}
img.ThrowIfDisposed();
descriptors.ThrowIfNotReady();
using (VectorOfKeyPoint keypointsVec = new VectorOfKeyPoint())
{
NativeMethods.nonfree_SURF_run2_OutputArray(ptr, img.CvPtr, Cv2.ToPtr(mask), keypointsVec.CvPtr,
descriptors.CvPtr, useProvidedKeypoints ? 1 : 0);
keypoints = keypointsVec.ToArray();
}
}
/// <summary>
/// Compute the BRISK features and descriptors on an image
/// </summary>
/// <param name="image"></param>
/// <param name="mask"></param>
/// <param name="keyPoints"></param>
/// <param name="descriptors"></param>
/// <param name="useProvidedKeypoints"></param>
public void Run(InputArray image, InputArray mask, out KeyPoint[] keyPoints,
OutputArray descriptors, bool useProvidedKeypoints = false)
{
ThrowIfDisposed();
if (image == null)
{
throw new ArgumentNullException("image");
}
if (descriptors == null)
{
throw new ArgumentNullException("descriptors");
}
image.ThrowIfDisposed();
descriptors.ThrowIfNotReady();
using (VectorOfKeyPoint keyPointsVec = new VectorOfKeyPoint())
{
NativeMethods.features2d_BRISK_run2(ptr, image.CvPtr, Cv2.ToPtr(mask), keyPointsVec.CvPtr,
descriptors.CvPtr, useProvidedKeypoints ? 1 : 0);
keyPoints = keyPointsVec.ToArray();
}
descriptors.Fix();
}
/// <summary>
///
/// </summary>
/// <param name="trainData"></param>
/// <param name="tflag"></param>
/// <param name="responses"></param>
/// <param name="varIdx"></param>
/// <param name="sampleIdx"></param>
/// <param name="varType"></param>
/// <param name="missingMask"></param>
/// <param name="param"></param>
/// <returns></returns>
#else
/// <summary>
///
/// </summary>
/// <param name="trainData"></param>
/// <param name="tflag"></param>
/// <param name="responses"></param>
/// <param name="varIdx"></param>
/// <param name="sampleIdx"></param>
/// <param name="varType"></param>
/// <param name="missingMask"></param>
/// <param name="param"></param>
/// <returns></returns>
#endif
public virtual bool Train(
CvMat trainData,
int tflag,
CvMat responses,
CvMat varIdx = null,
CvMat sampleIdx = null,
CvMat varType = null,
CvMat missingMask = null,
CvRTParams param = null)
{
if (trainData == null)
{
throw new ArgumentNullException("trainData");
}
if (responses == null)
{
throw new ArgumentNullException("responses");
}
trainData.ThrowIfDisposed();
responses.ThrowIfDisposed();
if (param == null)
{
param = new CvRTParams();
}
return(NativeMethods.ml_CvERTrees_train1(
ptr,
trainData.CvPtr,
tflag,
responses.CvPtr,
Cv2.ToPtr(varIdx),
Cv2.ToPtr(sampleIdx),
Cv2.ToPtr(varType),
Cv2.ToPtr(missingMask),
param.CvPtr) != 0);
}
/// <summary>
/// 決定木を学習する
/// </summary>
/// <param name="trainData"></param>
/// <param name="tflag"></param>
/// <param name="responses"></param>
/// <param name="varIdx"></param>
/// <param name="sampleIdx"></param>
/// <param name="varType"></param>
/// <param name="missingMask"></param>
/// <param name="param"></param>
/// <returns></returns>
#else
/// <summary>
/// Trains decision tree
/// </summary>
/// <param name="trainData"></param>
/// <param name="tflag"></param>
/// <param name="responses"></param>
/// <param name="varIdx"></param>
/// <param name="sampleIdx"></param>
/// <param name="varType"></param>
/// <param name="missingMask"></param>
/// <param name="param"></param>
/// <returns></returns>
#endif
public virtual bool Train(
Mat trainData,
DTreeDataLayout tflag,
Mat responses,
Mat varIdx,
Mat sampleIdx,
Mat varType,
Mat missingMask,
CvDTreeParams param)
{
if (trainData == null)
{
throw new ArgumentNullException(nameof(trainData));
}
if (responses == null)
{
throw new ArgumentNullException(nameof(responses));
}
trainData.ThrowIfDisposed();
responses.ThrowIfDisposed();
if (param == null)
{
param = new CvDTreeParams();
}
return(NativeMethods.ml_CvDTree_train_Mat(
ptr,
trainData.CvPtr,
(int)tflag,
responses.CvPtr,
Cv2.ToPtr(varIdx),
Cv2.ToPtr(sampleIdx),
Cv2.ToPtr(varType),
Cv2.ToPtr(missingMask),
param.CvPtr) != 0);
}
/// <summary>
/// 学習データを与えて初期化
/// </summary>
/// <param name="trainData"></param>
/// <param name="tflag"></param>
/// <param name="responses"></param>
/// <param name="varIdx"></param>
/// <param name="sampleIdx"></param>
/// <param name="varType"></param>
/// <param name="missingMask"></param>
/// <param name="param"></param>
#else
/// <summary>
/// Training constructor
/// </summary>
/// <param name="trainData"></param>
/// <param name="tflag"></param>
/// <param name="responses"></param>
/// <param name="varIdx"></param>
/// <param name="sampleIdx"></param>
/// <param name="varType"></param>
/// <param name="missingMask"></param>
/// <param name="param"></param>
#endif
public CvBoost(
CvMat trainData,
DTreeDataLayout tflag,
CvMat responses,
CvMat varIdx = null,
CvMat sampleIdx = null,
CvMat varType = null,
CvMat missingMask = null,
CvBoostParams param = null)
{
if (trainData == null)
{
throw new ArgumentNullException("trainData");
}
if (responses == null)
{
throw new ArgumentNullException("responses");
}
trainData.ThrowIfDisposed();
responses.ThrowIfDisposed();
if (param == null)
{
param = new CvBoostParams();
}
ptr = NativeMethods.ml_CvBoost_new_CvMat(
trainData.CvPtr,
(int)tflag,
responses.CvPtr,
Cv2.ToPtr(varIdx),
Cv2.ToPtr(sampleIdx),
Cv2.ToPtr(varType),
Cv2.ToPtr(missingMask),
param.CvPtr
);
}
/// <summary>
/// 入力サンプルに対する応答を予測する
/// </summary>
/// <param name="sample">入力サンプル</param>
/// <param name="missing">データ欠損マスク（オプション）．データ欠損を扱うためには，弱い決定木が代理分岐を含まなければならない．</param>
/// <param name="weakResponses">個々の弱い決定木からの応答の出力パラメータ（オプション）で，これは浮動小数点型ベクトルである．ベクトルの要素数は，slice 長と等しくなければならない．</param>
/// <param name="slice">予測に用いられる弱い決定木シーケンスの連続的部分集合（スライス）．デフォルトでは，全ての弱い分類器が用いられる．</param>
/// <param name="rawMode">falseは通常の入力を意味する．true の場合，このメソッドは離散入力変数の全ての値があらかじめ， 0..<num_of_categories_i>-1 の範囲に正規化されていることを仮定する（決定木は内部的にはこのような正規化された表現を用いている）．これは決定木集合の高速な予測に役立つ．連続変数の入力変数に対しては，このフラグは利用されない．</param>
/// <param name="returnSum"></param>
/// <returns>重み付き投票に基づく出力クラスラベル</returns>
#else
/// <summary>
/// Predicts response for the input sample
/// </summary>
/// <param name="sample">The input sample. </param>
/// <param name="missing">The optional mask of missing measurements. To handle missing measurements, the weak classifiers must include surrogate splits. </param>
/// <param name="weakResponses">The optional output parameter, a floating-point vector, of responses from each individual weak classifier. The number of elements in the vector must be equal to the slice length. </param>
/// <param name="slice">The continuous subset of the sequence of weak classifiers to be used for prediction. By default, all the weak classifiers are used. </param>
/// <param name="rawMode">The last parameter is normally set to false that implies a regular input. If it is true, the method assumes that all the values of the discrete input variables have been already normalized to 0..<num_of_categoriesi>-1 ranges. (as the decision tree uses such normalized representation internally). It is useful for faster prediction with tree ensembles. For ordered input variables the flag is not used. </param>
/// <param name="returnSum"></param>
/// <returns>the output class label based on the weighted voting. </returns>
#endif
public float Predict(
CvMat sample,
CvMat missing = null,
CvMat weakResponses = null,
CvSlice?slice = null,
bool rawMode = false,
bool returnSum = false)
{
if (sample == null)
{
throw new ArgumentNullException(nameof(sample));
}
CvSlice slice0 = slice.GetValueOrDefault(CvSlice.WholeSeq);
return(NativeMethods.ml_Boost_predict_CvMat(
ptr,
sample.CvPtr,
Cv2.ToPtr(missing),
Cv2.ToPtr(weakResponses),
slice0,
rawMode ? 1 : 0,
returnSum ? 1 : 0));
}
/// <summary>
/// モデルの学習
/// </summary>
/// <param name="trainData">既知のサンプル (m*n)</param>
/// <param name="responses">既知のサンプルのクラス (m*1)</param>
/// <param name="varIdx"></param>
/// <param name="sampleIdx"></param>
/// <param name="update">モデルを最初から学習する（false）か，新しい学習データを用いて更新する（true）か</param>
/// <returns></returns>
#else
/// <summary>
/// Trains the model
/// </summary>
/// <param name="trainData">Known samples (m*n)</param>
/// <param name="responses">Classes for known samples (m*1)</param>
/// <param name="varIdx"></param>
/// <param name="sampleIdx"></param>
/// <param name="update">Adds known samples to model(true) or makes a new one(false)</param>
/// <returns></returns>
#endif
public virtual bool Train(
Mat trainData,
Mat responses,
Mat varIdx,
Mat sampleIdx,
bool update = false)
{
if (trainData == null)
{
throw new ArgumentNullException("trainData");
}
if (responses == null)
{
throw new ArgumentNullException("responses");
}
return(NativeMethods.ml_CvNormalBayesClassifier_train_Mat(
ptr,
trainData.CvPtr,
responses.CvPtr,
Cv2.ToPtr(varIdx),
Cv2.ToPtr(sampleIdx),
update ? 1 : 0) != 0);
}
/// <summary>
/// 入力ベクトルの近傍を探す
/// </summary>
/// <param name="samples">既知のサンプル (l*n)</param>
/// <param name="k">探索する近傍の数の最大数</param>
/// <param name="results"></param>
/// <param name="neighbors"></param>
/// <param name="neighborResponses">それぞれのサンプルの近傍 (l*k)</param>
/// <param name="dist">サンプルから近傍までの距離</param>
/// <returns></returns>
#else
/// <summary>
/// Finds the K nearest neighbors of samples
/// </summary>
/// <param name="samples">Known samples (l*n)</param>
/// <param name="k">max neighbors to find</param>
/// <param name="results"></param>
/// <param name="neighbors"></param>
/// <param name="neighborResponses">Neighbors for each samples (l*k)</param>
/// <param name="dist">Distance from each sample to neighbors</param>
/// <returns></returns>
#endif
public virtual float FindNearest(
CvMat samples,
int k,
CvMat results = null,
float[][] neighbors = null,
CvMat neighborResponses = null,
CvMat dist = null)
{
if (samples == null)
{
throw new ArgumentNullException(nameof(samples));
}
if (neighbors == null)
{
return(NativeMethods.ml_CvKNearest_find_nearest_CvMat(
ptr,
samples.CvPtr,
k,
Cv2.ToPtr(results),
null,
Cv2.ToPtr(neighborResponses),
Cv2.ToPtr(dist)));
}
using (var aa = new ArrayAddress2 <Single>(neighbors))
{
return(NativeMethods.ml_CvKNearest_find_nearest_CvMat(
ptr,
samples.CvPtr,
k,
Cv2.ToPtr(results),
aa.Pointer,
Cv2.ToPtr(neighborResponses),
Cv2.ToPtr(dist)));
}
}
/// <summary>
/// サンプル集合からガウス混合パラメータを推定する
/// </summary>
/// <param name="samples"></param>
/// <param name="means0"></param>
/// <param name="covs0"></param>
/// <param name="weights0"></param>
/// <param name="logLikelihoods"></param>
/// <param name="labels"></param>
/// <param name="probs"></param>
#else
/// <summary>
/// Estimates Gaussian mixture parameters from the sample set
/// </summary>
/// <param name="samples"></param>
/// <param name="means0"></param>
/// <param name="covs0"></param>
/// <param name="weights0"></param>
/// <param name="logLikelihoods"></param>
/// <param name="labels"></param>
/// <param name="probs"></param>
#endif
public virtual bool TrainE(
InputArray samples,
InputArray means0,
InputArray covs0 = null,
InputArray weights0 = null,
OutputArray logLikelihoods = null,
OutputArray labels = null,
OutputArray probs = null)
{
if (disposed)
{
throw new ObjectDisposedException("EM");
}
if (samples == null)
{
throw new ArgumentNullException(nameof(samples));
}
if (means0 == null)
{
throw new ArgumentNullException(nameof(means0));
}
samples.ThrowIfDisposed();
means0.ThrowIfDisposed();
if (logLikelihoods != null)
{
logLikelihoods.ThrowIfNotReady();
}
if (covs0 != null)
{
covs0.ThrowIfDisposed();
}
if (weights0 != null)
{
weights0.ThrowIfDisposed();
}
if (labels != null)
{
labels.ThrowIfNotReady();
}
if (probs != null)
{
probs.ThrowIfNotReady();
}
int ret = NativeMethods.ml_EM_trainE(
ptr,
samples.CvPtr,
means0.CvPtr,
Cv2.ToPtr(covs0),
Cv2.ToPtr(weights0),
Cv2.ToPtr(logLikelihoods),
Cv2.ToPtr(labels),
Cv2.ToPtr(probs));
if (logLikelihoods != null)
{
logLikelihoods.Fix();
}
if (labels != null)
{
labels.Fix();
}
if (probs != null)
{
probs.Fix();
}
return(ret != 0);
}
/// <summary>
///
/// </summary>
/// <param name="indexParams"></param>
/// <param name="searchParams"></param>
public FlannBasedMatcher(IndexParams indexParams = null, SearchParams searchParams = null)
{
ptr = NativeMethods.features2d_FlannBasedMatcher_new(
Cv2.ToPtr(indexParams), Cv2.ToPtr(searchParams));
}
/// <summary>
/// Computes an image descriptor using the set visual vocabulary.
/// </summary>
/// <param name="image">Image, for which the descriptor is computed.</param>
/// <param name="keypoints">Keypoints detected in the input image.</param>
/// <param name="imgDescriptor">Computed output image descriptor.</param>
/// <param name="pointIdxsOfClusters">pointIdxsOfClusters Indices of keypoints that belong to the cluster.
/// This means that pointIdxsOfClusters[i] are keypoint indices that belong to the i -th cluster(word of vocabulary) returned if it is non-zero.</param>
/// <param name="descriptors">Descriptors of the image keypoints that are returned if they are non-zero.</param>
public void Compute(Mat image, out KeyPoint[] keypoints, Mat imgDescriptor,
out int[][] pointIdxsOfClusters, Mat descriptors = null)
{
if (IsDisposed)
{
throw new ObjectDisposedException(GetType().Name);
}
if (image == null)
{
throw new ArgumentNullException(nameof(image));
}
if (imgDescriptor == null)
{
throw new ArgumentNullException(nameof(imgDescriptor));
}
using (var keypointsVec = new VectorOfKeyPoint())
using (var pointIdxsOfClustersVec = new VectorOfVectorInt())
{
NativeMethods.features2d_BOWImgDescriptorExtractor_compute1(ptr, image.CvPtr, keypointsVec.CvPtr,
imgDescriptor.CvPtr, pointIdxsOfClustersVec.CvPtr, Cv2.ToPtr(descriptors));
keypoints = keypointsVec.ToArray();
pointIdxsOfClusters = pointIdxsOfClustersVec.ToArray();
}
GC.KeepAlive(image);
GC.KeepAlive(imgDescriptor);
GC.KeepAlive(descriptors);
}
/// <summary>
/// Detect keypoints in an image.
/// </summary>
/// <param name="image">The image.</param>
/// <param name="mask">Mask specifying where to look for keypoints (optional).
/// Must be a char matrix with non-zero values in the region of interest.</param>
/// <returns>The detected keypoints.</returns>
public KeyPoint[] Detect(Mat image, Mat mask = null)
{
if (image == null)
{
throw new ArgumentNullException("image");
}
using (var keypoints = new VectorOfKeyPoint())
{
NativeMethods.features2d_FeatureDetector_detect(ptr, image.CvPtr, keypoints.CvPtr, Cv2.ToPtr(mask));
return(keypoints.ToArray());
}
}
