int extract_features(Mat image)
{
if(train_entity.size() == 0)
{
// load_sift_models();
load_dictionary();
load_SVM();
}
Mat input_image = image ;
std::vector<KeyPoint> extracted_keypoints = extract_sift_keypoints(input_image);
Mat extracted_descriptor = compute_descriptors(input_image, extracted_keypoints);
image *= 255;
Mat gray_channel(image.size(), CV_8UC1) ;
image.convertTo(gray_channel,CV_8UC1) ;
Mat sub_mat = Mat::ones(gray_channel.size(), gray_channel.type())*255;
subtract(sub_mat, gray_channel, gray_channel);
std::vector<float> hog_ders = extract_hog_features(gray_channel);
cout << "calculating feature vector..." << endl ;
Mat feature_vector = extract_features_mat(extracted_descriptor, hog_ders);
float predicted_label = svm.predict(feature_vector);
cout << "PREDICTED LABEL IS: " << predicted_label << endl;
return predicted_label;
}
void GpuSurfDetectorInternal::computeDescriptors()
{
int nFeatures = m_features.ftCount();
if(nFeatures > 0)
{
compute_descriptors(m_features.deviceDescriptors(), m_features.deviceFeatures(), m_features.ftCount());
ASRL_CHECK_CUDA_ERROR_DBG("compute descriptors");
}
}
/**
Finds SIFT features in an image using user-specified parameter values. All
detected features are stored in the array pointed to by \a feat.
@param img the image in which to detect features
@param fea a pointer to an array in which to store detected features
@param intvls the number of intervals sampled per octave of scale space
@param sigma the amount of Gaussian smoothing applied to each image level
before building the scale space representation for an octave
@param cont_thr a threshold on the value of the scale space function
\f$\left|D(\hat{x})\right|\f$, where \f$\hat{x}\f$ is a vector specifying
feature location and scale, used to reject unstable features; assumes
pixel values in the range [0, 1]
@param curv_thr threshold on a feature's ratio of principle curvatures
used to reject features that are too edge-like
@param img_dbl should be 1 if image doubling prior to scale space
construction is desired or 0 if not
@param descr_width the width, \f$n\f$, of the \f$n \times n\f$ array of
orientation histograms used to compute a feature's descriptor
@param descr_hist_bins the number of orientations in each of the
histograms in the array used to compute a feature's descriptor
@return Returns the number of keypoints stored in \a feat or -1 on failure
@see sift_keypoints()
*/
int _sift_features( IplImage* img, struct feature** feat, int intvls,
double sigma, double contr_thr, int curv_thr,
int img_dbl, int descr_width, int descr_hist_bins )
{
IplImage* init_img;
IplImage*** gauss_pyr,***dog_pyr;
CvMemStorage* storage;
CvSeq* features;
int octvs, i, n = 0;
/* check arguments */
if( ! img )
fatal_error( "NULL pointer error, %s, line %d", __FILE__, __LINE__ );
if( ! feat )
fatal_error( "NULL pointer error, %s, line %d", __FILE__, __LINE__ );
/* build scale space pyramid; smallest dimension of top level is ~4 pixels */
init_img = create_init_img( img, img_dbl, sigma );
octvs = log( MIN( init_img->width, init_img->height ) ) / log(2) - 2;
gauss_pyr = build_gauss_pyr( init_img, octvs, intvls, sigma );
dog_pyr = build_dog_pyr( gauss_pyr, octvs, intvls );
storage = cvCreateMemStorage( 0 );
//极值检测
features = scale_space_extrema( dog_pyr, octvs, intvls, contr_thr,
curv_thr, storage );
//计算特征的尺度
calc_feature_scales( features, sigma, intvls );
//如果最先开始图像的尺寸被加倍了，那么调整features的参数
if( img_dbl )
adjust_for_img_dbl( features );
//计算特征点方向和幅值
calc_feature_oris( features, gauss_pyr );
//计算特征描述子,最难理解的函数
compute_descriptors( features, gauss_pyr, descr_width, descr_hist_bins );
/* sort features by decreasing scale and move from CvSeq to array */
cvSeqSort( features, (CvCmpFunc)feature_cmp, NULL );
n = features->total;
*feat = calloc( n, sizeof(struct feature) );
*feat = cvCvtSeqToArray( features, *feat, CV_WHOLE_SEQ );//features 里面放的是feature* 数据，所以*feat而不是feat
for( i = 0; i < n; i++ )//释放feature结构体中的feature_data，因为已经计算完了，得到了描述子
{
free( (*feat)[i].feature_data );
(*feat)[i].feature_data = NULL;
}
cvReleaseMemStorage( &storage );
cvReleaseImage( &init_img );
release_pyr( &gauss_pyr, octvs, intvls + 3 );
release_pyr( &dog_pyr, octvs, intvls + 2 );
return n;
}
/**
Finds SIFT features in an image using user-specified parameter values. All
detected features are stored in the array pointed to by \a feat.
@param img the image in which to detect features
@param fea a pointer to an array in which to store detected features
@param intvls the number of intervals sampled per octave of scale space
@param sigma the amount of Gaussian smoothing applied to each image level
before building the scale space representation for an octave
@param cont_thr a threshold on the value of the scale space function
\f$\left|D(\hat{x})\right|\f$, where \f$\hat{x}\f$ is a vector specifying
feature location and scale, used to reject unstable features; assumes
pixel values in the range [0, 1]
@param curv_thr threshold on a feature's ratio of principle curvatures
used to reject features that are too edge-like
@param img_dbl should be 1 if image doubling prior to scale space
construction is desired or 0 if not
@param descr_width the width, \f$n\f$, of the \f$n \times n\f$ array of
orientation histograms used to compute a feature's descriptor
@param descr_hist_bins the number of orientations in each of the
histograms in the array used to compute a feature's descriptor
@return Returns the number of keypoints stored in \a feat or -1 on failure
@see sift_keypoints()
*/
int _sift_features( IplImage* img, struct feature** feat, int intvls,
double sigma, double contr_thr, int curv_thr,
int img_dbl, int descr_width, int descr_hist_bins )
{
IplImage *init_img;
IplImage ***gauss_pyr, ***dog_pyr;
CvMemStorage *storage;
CvSeq *features;
int octvs, i, n = 0;
/* build scale space pyramid; smallest dimension of top level is ~4 pixels */
init_img = create_init_img( img, img_dbl, sigma );
octvs = log( MIN( init_img->width, init_img->height ) ) / log(2) - 2;
gauss_pyr = build_gauss_pyr( init_img, octvs, intvls, sigma );
dog_pyr = build_dog_pyr( gauss_pyr, octvs, intvls );
storage = cvCreateMemStorage( 0 );
features = scale_space_extrema( dog_pyr, octvs, intvls, contr_thr, curv_thr, storage );
calc_feature_scales( features, sigma, intvls );
if( img_dbl )
adjust_for_img_dbl( features );
calc_feature_oris( features, gauss_pyr );
compute_descriptors( features, gauss_pyr, descr_width, descr_hist_bins );
/* sort features by decreasing scale and move from CvSeq to array */
cvSeqSort( features, (CvCmpFunc)feature_cmp, NULL );
n = features->total;
*feat = calloc( n, sizeof(struct feature) );
*feat = cvCvtSeqToArray( features, *feat, CV_WHOLE_SEQ );
for( i = 0; i < n; i++ ) {
free( (*feat)[i].feature_data );
(*feat)[i].feature_data = NULL;
}
cvReleaseMemStorage( &storage );
cvReleaseImage( &init_img );
release_pyr( &gauss_pyr, octvs, intvls + 3 );
release_pyr( &dog_pyr, octvs, intvls + 2 );
return n;
}
/**
Finds SIFT features in an image using user-specified parameter values. All
detected features are stored in the array pointed to by \a feat.
@param img the image in which to detect features
@param feat a pointer to an array in which to store detected features
@param intvls the number of intervals sampled per octave of scale space
@param sigma the amount of Gaussian smoothing applied to each image level
before building the scale space representation for an octave
@param cont_thr a threshold on the value of the scale space function
\f$\left|D(\hat{x})\right|\f$, where \f$\hat{x}\f$ is a vector specifying
feature location and scale, used to reject unstable features; assumes
pixel values in the range [0, 1]
@param curv_thr threshold on a feature's ratio of principle curvatures
used to reject features that are too edge-like
@param img_dbl should be 1 if image doubling prior to scale space
construction is desired or 0 if not
@param descr_width the width, \f$n\f$, of the \f$n \times n\f$ array of
orientation histograms used to compute a feature's descriptor
@param descr_hist_bins the number of orientations in each of the
histograms in the array used to compute a feature's descriptor
@return Returns the number of keypoints stored in \a feat or -1 on failure
@see sift_keypoints()
*/
int _sift_features( IplImage* img, struct feature** feat, int intvls,
double sigma, double contr_thr, int curv_thr,
int img_dbl, int descr_width, int descr_hist_bins )
{
IplImage* init_img;//原图经初始化后的图像，归一化的32位灰度图
IplImage*** gauss_pyr, *** dog_pyr;//三级指针，高斯金字塔图像组，DoG金字塔图像组
CvMemStorage* storage;//存储器
CvSeq* features;//存储特征点的序列，序列中存放的是struct feature类型的指针
int octvs, i, n = 0;
//输入参数检查
/* check arguments */
if( ! img )
fatal_error( "NULL pointer error, %s, line %d", __FILE__, __LINE__ );
if( ! feat )
fatal_error( "NULL pointer error, %s, line %d", __FILE__, __LINE__ );
/* build scale space pyramid; smallest dimension of top level is ~4 pixels */
//██步骤一：██：建立尺度空间，即建立高斯差分(DoG)金字塔dog_pyr
//将原图转换为32位灰度图并归一化，然后进行一次高斯平滑，并根据参数img_dbl决定是否将图像尺寸放大为原图的2倍
init_img = create_init_img( img, img_dbl, sigma );
//计算高斯金字塔的组数octvs
octvs = log( MIN( init_img->width, init_img->height ) ) / log(2) - 2;
//为了保证连续性，在每一层的顶层继续用高斯模糊生成3幅图像，所以高斯金字塔每组有intvls+3层，DOG金字塔每组有intvls+2层
//建立高斯金字塔gauss_pyr，是一个octvs*(intvls+3)的图像数组
gauss_pyr = build_gauss_pyr( init_img, octvs, intvls, sigma );
//建立高斯差分(DoG)金字塔dog_pyr，是一个octvs*(intvls+2)的图像数组
dog_pyr = build_dog_pyr( gauss_pyr, octvs, intvls );
//██步骤二：██：在尺度空间中检测极值点，并进行精确定位和筛选
//创建默认大小的内存存储器
storage = cvCreateMemStorage( 0 );
//在尺度空间中检测极值点，通过插值精确定位，去除低对比度的点，去除边缘点，返回检测到的特征点序列
features = scale_space_extrema( dog_pyr, octvs, intvls, contr_thr, curv_thr, storage );
//计算特征点序列features中每个特征点的尺度
calc_feature_scales( features, sigma, intvls );
//若设置了将图像放大为原图的2倍
if( img_dbl )//将特征点序列中每个特征点的坐标减半(当设置了将图像放大为原图的2倍时，特征点检测完之后调用)
adjust_for_img_dbl( features );
//██步骤三：██：特征点方向赋值，完成此步骤后，每个特征点有三个信息：位置、尺度、方向
//计算每个特征点的梯度直方图，找出其主方向，若一个特征点有不止一个主方向，将其分为两个特征点
calc_feature_oris( features, gauss_pyr );
//██步骤四：██：计算特征描述子
//计算特征点序列中每个特征点的特征描述子向量
compute_descriptors( features, gauss_pyr, descr_width, descr_hist_bins );
/* sort features by decreasing scale and move from CvSeq to array */
//按特征点尺度的降序排列序列中元素的顺序，feature_cmp是自定义的比较函数
cvSeqSort( features, (CvCmpFunc)feature_cmp, NULL );
//将CvSeq类型的特征点序列features转换为通用的struct feature类型的数组feat
n = features->total;//特征点个数
*feat = calloc( n, sizeof(struct feature) );//分配控件
//将序列features中的元素拷贝到数组feat中，返回数组指针给feat
*feat = cvCvtSeqToArray( features, *feat, CV_WHOLE_SEQ );
//释放特征点数组feat中所有特征点的feature_data成员，因为此成员中的数据在检测完特征点后就没用了
for( i = 0; i < n; i++ )
{
free( (*feat)[i].feature_data );
(*feat)[i].feature_data = NULL;
}
//释放各种临时数据的存储空间
cvReleaseMemStorage( &storage );//释放内存存储器
cvReleaseImage( &init_img );//释放初始化后的图像
release_pyr( &gauss_pyr, octvs, intvls + 3 );//释放高斯金字塔图像组
release_pyr( &dog_pyr, octvs, intvls + 2 );//释放高斯差分金字塔图像组
return n;//返回检测到的特征点的个数
}