generic-library/opencv
1
0
Fork 0
Code Issues Pull Requests Releases Wiki Activity

Wrapped nldiffusion functions with details::kaze or details::amaze namespace to avoid collision of function names

Browse Source
This commit is contained in:
Ievgen Khvedchenia 2014-04-28 20:41:44 +03:00
parent c68cbfced3
commit c1bf453266
7 changed files with 697 additions and 681 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

2
modules/features2d/src/akaze/AKAZEConfig.h
View File

@ -114,7 +114,7 @@ struct AKAZEOptions {
float kcontrast; ///< The contrast factor parameter float kcontrast; ///< The contrast factor parameter
float kcontrast_percentile; ///< Percentile level for the contrast factor float kcontrast_percentile; ///< Percentile level for the contrast factor
size_t kcontrast_nbins; ///< Number of bins for the contrast factor histogram int kcontrast_nbins; ///< Number of bins for the contrast factor histogram
bool save_scale_space; ///< Set to true for saving the scale space images bool save_scale_space; ///< Set to true for saving the scale space images
bool save_keypoints; ///< Set to true for saving the detected keypoints and descriptors bool save_keypoints; ///< Set to true for saving the detected keypoints and descriptors

4
modules/features2d/src/akaze/AKAZEFeatures.cpp
View File

@ -12,6 +12,7 @@
using namespace std; using namespace std;
using namespace cv; using namespace cv;
using namespace cv::details::akaze;
/* ************************************************************************* */ /* ************************************************************************* */
/** /**
@ -110,8 +111,7 @@ int AKAZEFeatures::Create_Nonlinear_Scale_Space(const cv::Mat& img) {
evolution_[0].Lt.copyTo(evolution_[0].Lsmooth); evolution_[0].Lt.copyTo(evolution_[0].Lsmooth);
// First compute the kcontrast factor // First compute the kcontrast factor
options_.kcontrast = compute_k_percentile(img, options_.kcontrast_percentile, options_.kcontrast = compute_k_percentile(img, options_.kcontrast_percentile, 1.0f, options_.kcontrast_nbins, 0, 0);
1.0f, options_.kcontrast_nbins, 0, 0);
//t2 = cv::getTickCount(); //t2 = cv::getTickCount();
//timing_.kcontrast = 1000.0*(t2 - t1) / cv::getTickFrequency(); //timing_.kcontrast = 1000.0*(t2 - t1) / cv::getTickFrequency();

689
modules/features2d/src/akaze/nldiffusion_functions.cpp
View File

@ -19,368 +19,373 @@
* @author Pablo F. Alcantarilla, Jesus Nuevo * @author Pablo F. Alcantarilla, Jesus Nuevo
*/ */
#include "nldiffusion_functions.h" #include "akaze/nldiffusion_functions.h"
using namespace std; using namespace std;
using namespace cv; using namespace cv;
/* ************************************************************************* */ namespace cv {
/** namespace details {
* @brief This function smoothes an image with a Gaussian kernel namespace akaze {
* @param src Input image
* @param dst Output image
* @param ksize_x Kernel size in X-direction (horizontal)
* @param ksize_y Kernel size in Y-direction (vertical)
* @param sigma Kernel standard deviation
*/
void gaussian_2D_convolution(const cv::Mat& src, cv::Mat& dst, const size_t& ksize_x,
const size_t& ksize_y, const float& sigma) {
int ksize_x_ = 0, ksize_y_ = 0; /* ************************************************************************* */
/**
* @brief This function smoothes an image with a Gaussian kernel
* @param src Input image
* @param dst Output image
* @param ksize_x Kernel size in X-direction (horizontal)
* @param ksize_y Kernel size in Y-direction (vertical)
* @param sigma Kernel standard deviation
*/
void gaussian_2D_convolution(const cv::Mat& src, cv::Mat& dst, int ksize_x, int ksize_y, float sigma) {
// Compute an appropriate kernel size according to the specified sigma int ksize_x_ = 0, ksize_y_ = 0;
if (sigma > ksize_x || sigma > ksize_y || ksize_x == 0 || ksize_y == 0) {
ksize_x_ = (int)ceil(2.0f*(1.0f + (sigma - 0.8f) / (0.3f)));
ksize_y_ = ksize_x_;
}
// The kernel size must be and odd number // Compute an appropriate kernel size according to the specified sigma
if ((ksize_x_ % 2) == 0) { if (sigma > ksize_x || sigma > ksize_y || ksize_x == 0 || ksize_y == 0) {
ksize_x_ += 1; ksize_x_ = (int)ceil(2.0f*(1.0f + (sigma - 0.8f) / (0.3f)));
} ksize_y_ = ksize_x_;
if ((ksize_y_ % 2) == 0) {
ksize_y_ += 1;
}
// Perform the Gaussian Smoothing with border replication
GaussianBlur(src, dst, Size(ksize_x_, ksize_y_), sigma, sigma, BORDER_REPLICATE);
}
/* ************************************************************************* */
/**
* @brief This function computes image derivatives with Scharr kernel
* @param src Input image
* @param dst Output image
* @param xorder Derivative order in X-direction (horizontal)
* @param yorder Derivative order in Y-direction (vertical)
* @note Scharr operator approximates better rotation invariance than
* other stencils such as Sobel. See Weickert and Scharr,
* A Scheme for Coherence-Enhancing Diffusion Filtering with Optimized Rotation Invariance,
* Journal of Visual Communication and Image Representation 2002
*/
void image_derivatives_scharr(const cv::Mat& src, cv::Mat& dst, int xorder, int yorder) {
Scharr(src, dst, CV_32F, xorder, yorder, 1.0, 0, BORDER_DEFAULT);
}
/* ************************************************************************* */
/**
* @brief This function computes the Perona and Malik conductivity coefficient g1
* g1 = exp(-|dL|^2/k^2)
* @param Lx First order image derivative in X-direction (horizontal)
* @param Ly First order image derivative in Y-direction (vertical)
* @param dst Output image
* @param k Contrast factor parameter
*/
void pm_g1(const cv::Mat& Lx, const cv::Mat& Ly, cv::Mat& dst, const float& k) {
exp(-(Lx.mul(Lx) + Ly.mul(Ly)) / (k*k), dst);
}
/* ************************************************************************* */
/**
* @brief This function computes the Perona and Malik conductivity coefficient g2
* g2 = 1 / (1 + dL^2 / k^2)
* @param Lx First order image derivative in X-direction (horizontal)
* @param Ly First order image derivative in Y-direction (vertical)
* @param dst Output image
* @param k Contrast factor parameter
*/
void pm_g2(const cv::Mat& Lx, const cv::Mat& Ly, cv::Mat& dst, const float& k) {
dst = 1.0 / (1.0 + (Lx.mul(Lx) + Ly.mul(Ly)) / (k*k));
}
/* ************************************************************************* */
/**
* @brief This function computes Weickert conductivity coefficient gw
* @param Lx First order image derivative in X-direction (horizontal)
* @param Ly First order image derivative in Y-direction (vertical)
* @param dst Output image
* @param k Contrast factor parameter
* @note For more information check the following paper: J. Weickert
* Applications of nonlinear diffusion in image processing and computer vision,
* Proceedings of Algorithmy 2000
*/
void weickert_diffusivity(const cv::Mat& Lx, const cv::Mat& Ly, cv::Mat& dst, const float& k) {
Mat modg;
pow((Lx.mul(Lx) + Ly.mul(Ly)) / (k*k), 4, modg);
cv::exp(-3.315 / modg, dst);
dst = 1.0 - dst;
}
/* ************************************************************************* */
/**
* @brief This function computes Charbonnier conductivity coefficient gc
* gc = 1 / sqrt(1 + dL^2 / k^2)
* @param Lx First order image derivative in X-direction (horizontal)
* @param Ly First order image derivative in Y-direction (vertical)
* @param dst Output image
* @param k Contrast factor parameter
* @note For more information check the following paper: J. Weickert
* Applications of nonlinear diffusion in image processing and computer vision,
* Proceedings of Algorithmy 2000
*/
void charbonnier_diffusivity(const cv::Mat& Lx, const cv::Mat& Ly, cv::Mat& dst, const float& k) {
Mat den;
cv::sqrt(1.0 + (Lx.mul(Lx) + Ly.mul(Ly)) / (k*k), den);
dst = 1.0 / den;
}
/* ************************************************************************* */
/**
* @brief This function computes a good empirical value for the k contrast factor
* given an input image, the percentile (0-1), the gradient scale and the number of
* bins in the histogram
* @param img Input image
* @param perc Percentile of the image gradient histogram (0-1)
* @param gscale Scale for computing the image gradient histogram
* @param nbins Number of histogram bins
* @param ksize_x Kernel size in X-direction (horizontal) for the Gaussian smoothing kernel
* @param ksize_y Kernel size in Y-direction (vertical) for the Gaussian smoothing kernel
* @return k contrast factor
*/
float compute_k_percentile(const cv::Mat& img, float perc, float gscale,
size_t nbins, size_t ksize_x, size_t ksize_y) {
size_t nbin = 0, nelements = 0, nthreshold = 0, k = 0;
float kperc = 0.0, modg = 0.0, lx = 0.0, ly = 0.0;
float npoints = 0.0;
float hmax = 0.0;
// Create the array for the histogram
std::vector<size_t> hist(nbins, 0);
// Create the matrices
cv::Mat gaussian = cv::Mat::zeros(img.rows, img.cols, CV_32F);
cv::Mat Lx = cv::Mat::zeros(img.rows, img.cols, CV_32F);
cv::Mat Ly = cv::Mat::zeros(img.rows, img.cols, CV_32F);
// Perform the Gaussian convolution
gaussian_2D_convolution(img, gaussian, ksize_x, ksize_y, gscale);
// Compute the Gaussian derivatives Lx and Ly
image_derivatives_scharr(gaussian, Lx, 1, 0);
image_derivatives_scharr(gaussian, Ly, 0, 1);
// Skip the borders for computing the histogram
for (int i = 1; i < gaussian.rows - 1; i++) {
for (int j = 1; j < gaussian.cols - 1; j++) {
lx = *(Lx.ptr<float>(i)+j);
ly = *(Ly.ptr<float>(i)+j);
modg = sqrt(lx*lx + ly*ly);
// Get the maximum
if (modg > hmax) {
hmax = modg;
}
}
}
// Skip the borders for computing the histogram
for (int i = 1; i < gaussian.rows - 1; i++) {
for (int j = 1; j < gaussian.cols - 1; j++) {
lx = *(Lx.ptr<float>(i)+j);
ly = *(Ly.ptr<float>(i)+j);
modg = sqrt(lx*lx + ly*ly);
// Find the correspondent bin
if (modg != 0.0) {
nbin = (size_t)floor(nbins*(modg / hmax));
if (nbin == nbins) {
nbin--;
} }
hist[nbin]++; // The kernel size must be and odd number
npoints++; if ((ksize_x_ % 2) == 0) {
ksize_x_ += 1;
}
if ((ksize_y_ % 2) == 0) {
ksize_y_ += 1;
}
// Perform the Gaussian Smoothing with border replication
GaussianBlur(src, dst, Size(ksize_x_, ksize_y_), sigma, sigma, BORDER_REPLICATE);
} }
}
}
// Now find the perc of the histogram percentile /* ************************************************************************* */
nthreshold = (size_t)(npoints*perc); /**
* @brief This function computes image derivatives with Scharr kernel
* @param src Input image
* @param dst Output image
* @param xorder Derivative order in X-direction (horizontal)
* @param yorder Derivative order in Y-direction (vertical)
* @note Scharr operator approximates better rotation invariance than
* other stencils such as Sobel. See Weickert and Scharr,
* A Scheme for Coherence-Enhancing Diffusion Filtering with Optimized Rotation Invariance,
* Journal of Visual Communication and Image Representation 2002
*/
void image_derivatives_scharr(const cv::Mat& src, cv::Mat& dst, int xorder, int yorder) {
Scharr(src, dst, CV_32F, xorder, yorder, 1.0, 0, BORDER_DEFAULT);
}
for (k = 0; nelements < nthreshold && k < nbins; k++) { /* ************************************************************************* */
nelements = nelements + hist[k]; /**
} * @brief This function computes the Perona and Malik conductivity coefficient g1
* g1 = exp(-|dL|^2/k^2)
* @param Lx First order image derivative in X-direction (horizontal)
* @param Ly First order image derivative in Y-direction (vertical)
* @param dst Output image
* @param k Contrast factor parameter
*/
void pm_g1(const cv::Mat& Lx, const cv::Mat& Ly, cv::Mat& dst, const float& k) {
exp(-(Lx.mul(Lx) + Ly.mul(Ly)) / (k*k), dst);
}
if (nelements < nthreshold) { /* ************************************************************************* */
kperc = 0.03f; /**
} * @brief This function computes the Perona and Malik conductivity coefficient g2
else { * g2 = 1 / (1 + dL^2 / k^2)
kperc = hmax*((float)(k) / (float)nbins); * @param Lx First order image derivative in X-direction (horizontal)
} * @param Ly First order image derivative in Y-direction (vertical)
* @param dst Output image
* @param k Contrast factor parameter
*/
void pm_g2(const cv::Mat& Lx, const cv::Mat& Ly, cv::Mat& dst, const float& k) {
dst = 1.0 / (1.0 + (Lx.mul(Lx) + Ly.mul(Ly)) / (k*k));
}
return kperc; /* ************************************************************************* */
} /**
* @brief This function computes Weickert conductivity coefficient gw
* @param Lx First order image derivative in X-direction (horizontal)
* @param Ly First order image derivative in Y-direction (vertical)
* @param dst Output image
* @param k Contrast factor parameter
* @note For more information check the following paper: J. Weickert
* Applications of nonlinear diffusion in image processing and computer vision,
* Proceedings of Algorithmy 2000
*/
void weickert_diffusivity(const cv::Mat& Lx, const cv::Mat& Ly, cv::Mat& dst, const float& k) {
Mat modg;
pow((Lx.mul(Lx) + Ly.mul(Ly)) / (k*k), 4, modg);
cv::exp(-3.315 / modg, dst);
dst = 1.0 - dst;
}
/* ************************************************************************* */ /* ************************************************************************* */
/** /**
* @brief This function computes Scharr image derivatives * @brief This function computes Charbonnier conductivity coefficient gc
* @param src Input image * gc = 1 / sqrt(1 + dL^2 / k^2)
* @param dst Output image * @param Lx First order image derivative in X-direction (horizontal)
* @param xorder Derivative order in X-direction (horizontal) * @param Ly First order image derivative in Y-direction (vertical)
* @param yorder Derivative order in Y-direction (vertical) * @param dst Output image
* @param scale Scale factor for the derivative size * @param k Contrast factor parameter
*/ * @note For more information check the following paper: J. Weickert
void compute_scharr_derivatives(const cv::Mat& src, cv::Mat& dst, int xorder, int yorder, int scale) { * Applications of nonlinear diffusion in image processing and computer vision,
* Proceedings of Algorithmy 2000
*/
void charbonnier_diffusivity(const cv::Mat& Lx, const cv::Mat& Ly, cv::Mat& dst, const float& k) {
Mat den;
cv::sqrt(1.0 + (Lx.mul(Lx) + Ly.mul(Ly)) / (k*k), den);
dst = 1.0 / den;
}
Mat kx, ky; /* ************************************************************************* */
compute_derivative_kernels(kx, ky, xorder, yorder, scale); /**
sepFilter2D(src, dst, CV_32F, kx, ky); * @brief This function computes a good empirical value for the k contrast factor
} * given an input image, the percentile (0-1), the gradient scale and the number of
* bins in the histogram
* @param img Input image
* @param perc Percentile of the image gradient histogram (0-1)
* @param gscale Scale for computing the image gradient histogram
* @param nbins Number of histogram bins
* @param ksize_x Kernel size in X-direction (horizontal) for the Gaussian smoothing kernel
* @param ksize_y Kernel size in Y-direction (vertical) for the Gaussian smoothing kernel
* @return k contrast factor
*/
float compute_k_percentile(const cv::Mat& img, float perc, float gscale, int nbins, int ksize_x, int ksize_y) {
/* ************************************************************************* */ int nbin = 0, nelements = 0, nthreshold = 0, k = 0;
/** float kperc = 0.0, modg = 0.0, lx = 0.0, ly = 0.0;
* @brief This function performs a scalar non-linear diffusion step float npoints = 0.0;
* @param Ld2 Output image in the evolution float hmax = 0.0;
* @param c Conductivity image
* @param Lstep Previous image in the evolution // Create the array for the histogram
* @param stepsize The step size in time units std::vector<int> hist(nbins, 0);
* @note Forward Euler Scheme 3x3 stencil
* The function c is a scalar value that depends on the gradient norm // Create the matrices
* dL_by_ds = d(c dL_by_dx)_by_dx + d(c dL_by_dy)_by_dy cv::Mat gaussian = cv::Mat::zeros(img.rows, img.cols, CV_32F);
*/ cv::Mat Lx = cv::Mat::zeros(img.rows, img.cols, CV_32F);
void nld_step_scalar(cv::Mat& Ld, const cv::Mat& c, cv::Mat& Lstep, const float& stepsize) { cv::Mat Ly = cv::Mat::zeros(img.rows, img.cols, CV_32F);
// Perform the Gaussian convolution
gaussian_2D_convolution(img, gaussian, ksize_x, ksize_y, gscale);
// Compute the Gaussian derivatives Lx and Ly
image_derivatives_scharr(gaussian, Lx, 1, 0);
image_derivatives_scharr(gaussian, Ly, 0, 1);
// Skip the borders for computing the histogram
for (int i = 1; i < gaussian.rows - 1; i++) {
for (int j = 1; j < gaussian.cols - 1; j++) {
lx = *(Lx.ptr<float>(i)+j);
ly = *(Ly.ptr<float>(i)+j);
modg = sqrt(lx*lx + ly*ly);
// Get the maximum
if (modg > hmax) {
hmax = modg;
}
}
}
// Skip the borders for computing the histogram
for (int i = 1; i < gaussian.rows - 1; i++) {
for (int j = 1; j < gaussian.cols - 1; j++) {
lx = *(Lx.ptr<float>(i)+j);
ly = *(Ly.ptr<float>(i)+j);
modg = sqrt(lx*lx + ly*ly);
// Find the correspondent bin
if (modg != 0.0) {
nbin = (int)floor(nbins*(modg / hmax));
if (nbin == nbins) {
nbin--;
}
hist[nbin]++;
npoints++;
}
}
}
// Now find the perc of the histogram percentile
nthreshold = (int)(npoints*perc);
for (k = 0; nelements < nthreshold && k < nbins; k++) {
nelements = nelements + hist[k];
}
if (nelements < nthreshold) {
kperc = 0.03f;
}
else {
kperc = hmax*((float)(k) / (float)nbins);
}
return kperc;
}
/* ************************************************************************* */
/**
* @brief This function computes Scharr image derivatives
* @param src Input image
* @param dst Output image
* @param xorder Derivative order in X-direction (horizontal)
* @param yorder Derivative order in Y-direction (vertical)
* @param scale Scale factor for the derivative size
*/
void compute_scharr_derivatives(const cv::Mat& src, cv::Mat& dst, int xorder, int yorder, int scale) {
Mat kx, ky;
compute_derivative_kernels(kx, ky, xorder, yorder, scale);
sepFilter2D(src, dst, CV_32F, kx, ky);
}
/* ************************************************************************* */
/**
* @brief This function performs a scalar non-linear diffusion step
* @param Ld2 Output image in the evolution
* @param c Conductivity image
* @param Lstep Previous image in the evolution
* @param stepsize The step size in time units
* @note Forward Euler Scheme 3x3 stencil
* The function c is a scalar value that depends on the gradient norm
* dL_by_ds = d(c dL_by_dx)_by_dx + d(c dL_by_dy)_by_dy
*/
void nld_step_scalar(cv::Mat& Ld, const cv::Mat& c, cv::Mat& Lstep, const float& stepsize) {
#ifdef _OPENMP #ifdef _OPENMP
#pragma omp parallel for schedule(dynamic) #pragma omp parallel for schedule(dynamic)
#endif #endif
for (int i = 1; i < Lstep.rows - 1; i++) { for (int i = 1; i < Lstep.rows - 1; i++) {
for (int j = 1; j < Lstep.cols - 1; j++) { for (int j = 1; j < Lstep.cols - 1; j++) {
float xpos = ((*(c.ptr<float>(i)+j)) + (*(c.ptr<float>(i)+j + 1)))*((*(Ld.ptr<float>(i)+j + 1)) - (*(Ld.ptr<float>(i)+j))); float xpos = ((*(c.ptr<float>(i)+j)) + (*(c.ptr<float>(i)+j + 1)))*((*(Ld.ptr<float>(i)+j + 1)) - (*(Ld.ptr<float>(i)+j)));
float xneg = ((*(c.ptr<float>(i)+j - 1)) + (*(c.ptr<float>(i)+j)))*((*(Ld.ptr<float>(i)+j)) - (*(Ld.ptr<float>(i)+j - 1))); float xneg = ((*(c.ptr<float>(i)+j - 1)) + (*(c.ptr<float>(i)+j)))*((*(Ld.ptr<float>(i)+j)) - (*(Ld.ptr<float>(i)+j - 1)));
float ypos = ((*(c.ptr<float>(i)+j)) + (*(c.ptr<float>(i + 1) + j)))*((*(Ld.ptr<float>(i + 1) + j)) - (*(Ld.ptr<float>(i)+j))); float ypos = ((*(c.ptr<float>(i)+j)) + (*(c.ptr<float>(i + 1) + j)))*((*(Ld.ptr<float>(i + 1) + j)) - (*(Ld.ptr<float>(i)+j)));
float yneg = ((*(c.ptr<float>(i - 1) + j)) + (*(c.ptr<float>(i)+j)))*((*(Ld.ptr<float>(i)+j)) - (*(Ld.ptr<float>(i - 1) + j))); float yneg = ((*(c.ptr<float>(i - 1) + j)) + (*(c.ptr<float>(i)+j)))*((*(Ld.ptr<float>(i)+j)) - (*(Ld.ptr<float>(i - 1) + j)));
*(Lstep.ptr<float>(i)+j) = 0.5f*stepsize*(xpos - xneg + ypos - yneg); *(Lstep.ptr<float>(i)+j) = 0.5f*stepsize*(xpos - xneg + ypos - yneg);
}
}
for (int j = 1; j < Lstep.cols - 1; j++) {
float xpos = ((*(c.ptr<float>(0) + j)) + (*(c.ptr<float>(0) + j + 1)))*((*(Ld.ptr<float>(0) + j + 1)) - (*(Ld.ptr<float>(0) + j)));
float xneg = ((*(c.ptr<float>(0) + j - 1)) + (*(c.ptr<float>(0) + j)))*((*(Ld.ptr<float>(0) + j)) - (*(Ld.ptr<float>(0) + j - 1)));
float ypos = ((*(c.ptr<float>(0) + j)) + (*(c.ptr<float>(1) + j)))*((*(Ld.ptr<float>(1) + j)) - (*(Ld.ptr<float>(0) + j)));
*(Lstep.ptr<float>(0) + j) = 0.5f*stepsize*(xpos - xneg + ypos);
}
for (int j = 1; j < Lstep.cols - 1; j++) {
float xpos = ((*(c.ptr<float>(Lstep.rows - 1) + j)) + (*(c.ptr<float>(Lstep.rows - 1) + j + 1)))*((*(Ld.ptr<float>(Lstep.rows - 1) + j + 1)) - (*(Ld.ptr<float>(Lstep.rows - 1) + j)));
float xneg = ((*(c.ptr<float>(Lstep.rows - 1) + j - 1)) + (*(c.ptr<float>(Lstep.rows - 1) + j)))*((*(Ld.ptr<float>(Lstep.rows - 1) + j)) - (*(Ld.ptr<float>(Lstep.rows - 1) + j - 1)));
float ypos = ((*(c.ptr<float>(Lstep.rows - 1) + j)) + (*(c.ptr<float>(Lstep.rows - 1) + j)))*((*(Ld.ptr<float>(Lstep.rows - 1) + j)) - (*(Ld.ptr<float>(Lstep.rows - 1) + j)));
float yneg = ((*(c.ptr<float>(Lstep.rows - 2) + j)) + (*(c.ptr<float>(Lstep.rows - 1) + j)))*((*(Ld.ptr<float>(Lstep.rows - 1) + j)) - (*(Ld.ptr<float>(Lstep.rows - 2) + j)));
*(Lstep.ptr<float>(Lstep.rows - 1) + j) = 0.5f*stepsize*(xpos - xneg + ypos - yneg);
}
for (int i = 1; i < Lstep.rows - 1; i++) {
float xpos = ((*(c.ptr<float>(i))) + (*(c.ptr<float>(i)+1)))*((*(Ld.ptr<float>(i)+1)) - (*(Ld.ptr<float>(i))));
float xneg = ((*(c.ptr<float>(i))) + (*(c.ptr<float>(i))))*((*(Ld.ptr<float>(i))) - (*(Ld.ptr<float>(i))));
float ypos = ((*(c.ptr<float>(i))) + (*(c.ptr<float>(i + 1))))*((*(Ld.ptr<float>(i + 1))) - (*(Ld.ptr<float>(i))));
float yneg = ((*(c.ptr<float>(i - 1))) + (*(c.ptr<float>(i))))*((*(Ld.ptr<float>(i))) - (*(Ld.ptr<float>(i - 1))));
*(Lstep.ptr<float>(i)) = 0.5f*stepsize*(xpos - xneg + ypos - yneg);
}
for (int i = 1; i < Lstep.rows - 1; i++) {
float xneg = ((*(c.ptr<float>(i)+Lstep.cols - 2)) + (*(c.ptr<float>(i)+Lstep.cols - 1)))*((*(Ld.ptr<float>(i)+Lstep.cols - 1)) - (*(Ld.ptr<float>(i)+Lstep.cols - 2)));
float ypos = ((*(c.ptr<float>(i)+Lstep.cols - 1)) + (*(c.ptr<float>(i + 1) + Lstep.cols - 1)))*((*(Ld.ptr<float>(i + 1) + Lstep.cols - 1)) - (*(Ld.ptr<float>(i)+Lstep.cols - 1)));
float yneg = ((*(c.ptr<float>(i - 1) + Lstep.cols - 1)) + (*(c.ptr<float>(i)+Lstep.cols - 1)))*((*(Ld.ptr<float>(i)+Lstep.cols - 1)) - (*(Ld.ptr<float>(i - 1) + Lstep.cols - 1)));
*(Lstep.ptr<float>(i)+Lstep.cols - 1) = 0.5f*stepsize*(-xneg + ypos - yneg);
}
Ld = Ld + Lstep;
}
/* ************************************************************************* */
/**
* @brief This function downsamples the input image with the kernel [1/4,1/2,1/4]
* @param img Input image to be downsampled
* @param dst Output image with half of the resolution of the input image
*/
void downsample_image(const cv::Mat& src, cv::Mat& dst) {
int i1 = 0, j1 = 0, i2 = 0, j2 = 0;
for (i1 = 1; i1 < src.rows; i1 += 2) {
j2 = 0;
for (j1 = 1; j1 < src.cols; j1 += 2) {
*(dst.ptr<float>(i2)+j2) = 0.5f*(*(src.ptr<float>(i1)+j1)) + 0.25f*(*(src.ptr<float>(i1)+j1 - 1) + *(src.ptr<float>(i1)+j1 + 1));
j2++;
}
i2++;
}
}
/* ************************************************************************* */
/**
* @brief This function downsamples the input image using OpenCV resize
* @param img Input image to be downsampled
* @param dst Output image with half of the resolution of the input image
*/
void halfsample_image(const cv::Mat& src, cv::Mat& dst) {
// Make sure the destination image is of the right size
CV_Assert(src.cols / 2 == dst.cols);
CV_Assert(src.rows / 2 == dst.rows);
resize(src, dst, dst.size(), 0, 0, cv::INTER_AREA);
}
/* ************************************************************************* */
/**
* @brief Compute Scharr derivative kernels for sizes different than 3
* @param kx_ The derivative kernel in x-direction
* @param ky_ The derivative kernel in y-direction
* @param dx The derivative order in x-direction
* @param dy The derivative order in y-direction
* @param scale The kernel size
*/
void compute_derivative_kernels(cv::OutputArray kx_, cv::OutputArray ky_, int dx, int dy, int scale) {
const int ksize = 3 + 2 * (scale - 1);
// The usual Scharr kernel
if (scale == 1) {
getDerivKernels(kx_, ky_, dx, dy, 0, true, CV_32F);
return;
}
kx_.create(ksize, 1, CV_32F, -1, true);
ky_.create(ksize, 1, CV_32F, -1, true);
Mat kx = kx_.getMat();
Mat ky = ky_.getMat();
float w = 10.0f / 3.0f;
float norm = 1.0f / (2.0f*scale*(w + 2.0f));
for (int k = 0; k < 2; k++) {
Mat* kernel = k == 0 ? &kx : &ky;
int order = k == 0 ? dx : dy;
float kerI[1000];
for (int t = 0; t < ksize; t++) {
kerI[t] = 0;
}
if (order == 0) {
kerI[0] = norm;
kerI[ksize / 2] = w*norm;
kerI[ksize - 1] = norm;
}
else if (order == 1) {
kerI[0] = -1;
kerI[ksize / 2] = 0;
kerI[ksize - 1] = 1;
}
Mat temp(kernel->rows, kernel->cols, CV_32F, &kerI[0]);
temp.copyTo(*kernel);
}
}
} }
} }
}
for (int j = 1; j < Lstep.cols - 1; j++) {
float xpos = ((*(c.ptr<float>(0) + j)) + (*(c.ptr<float>(0) + j + 1)))*((*(Ld.ptr<float>(0) + j + 1)) - (*(Ld.ptr<float>(0) + j)));
float xneg = ((*(c.ptr<float>(0) + j - 1)) + (*(c.ptr<float>(0) + j)))*((*(Ld.ptr<float>(0) + j)) - (*(Ld.ptr<float>(0) + j - 1)));
float ypos = ((*(c.ptr<float>(0) + j)) + (*(c.ptr<float>(1) + j)))*((*(Ld.ptr<float>(1) + j)) - (*(Ld.ptr<float>(0) + j)));
*(Lstep.ptr<float>(0) + j) = 0.5f*stepsize*(xpos - xneg + ypos);
}
for (int j = 1; j < Lstep.cols - 1; j++) {
float xpos = ((*(c.ptr<float>(Lstep.rows - 1) + j)) + (*(c.ptr<float>(Lstep.rows - 1) + j + 1)))*((*(Ld.ptr<float>(Lstep.rows - 1) + j + 1)) - (*(Ld.ptr<float>(Lstep.rows - 1) + j)));
float xneg = ((*(c.ptr<float>(Lstep.rows - 1) + j - 1)) + (*(c.ptr<float>(Lstep.rows - 1) + j)))*((*(Ld.ptr<float>(Lstep.rows - 1) + j)) - (*(Ld.ptr<float>(Lstep.rows - 1) + j - 1)));
float ypos = ((*(c.ptr<float>(Lstep.rows - 1) + j)) + (*(c.ptr<float>(Lstep.rows - 1) + j)))*((*(Ld.ptr<float>(Lstep.rows - 1) + j)) - (*(Ld.ptr<float>(Lstep.rows - 1) + j)));
float yneg = ((*(c.ptr<float>(Lstep.rows - 2) + j)) + (*(c.ptr<float>(Lstep.rows - 1) + j)))*((*(Ld.ptr<float>(Lstep.rows - 1) + j)) - (*(Ld.ptr<float>(Lstep.rows - 2) + j)));
*(Lstep.ptr<float>(Lstep.rows - 1) + j) = 0.5f*stepsize*(xpos - xneg + ypos - yneg);
}
for (int i = 1; i < Lstep.rows - 1; i++) {
float xpos = ((*(c.ptr<float>(i))) + (*(c.ptr<float>(i)+1)))*((*(Ld.ptr<float>(i)+1)) - (*(Ld.ptr<float>(i))));
float xneg = ((*(c.ptr<float>(i))) + (*(c.ptr<float>(i))))*((*(Ld.ptr<float>(i))) - (*(Ld.ptr<float>(i))));
float ypos = ((*(c.ptr<float>(i))) + (*(c.ptr<float>(i + 1))))*((*(Ld.ptr<float>(i + 1))) - (*(Ld.ptr<float>(i))));
float yneg = ((*(c.ptr<float>(i - 1))) + (*(c.ptr<float>(i))))*((*(Ld.ptr<float>(i))) - (*(Ld.ptr<float>(i - 1))));
*(Lstep.ptr<float>(i)) = 0.5f*stepsize*(xpos - xneg + ypos - yneg);
}
for (int i = 1; i < Lstep.rows - 1; i++) {
float xneg = ((*(c.ptr<float>(i)+Lstep.cols - 2)) + (*(c.ptr<float>(i)+Lstep.cols - 1)))*((*(Ld.ptr<float>(i)+Lstep.cols - 1)) - (*(Ld.ptr<float>(i)+Lstep.cols - 2)));
float ypos = ((*(c.ptr<float>(i)+Lstep.cols - 1)) + (*(c.ptr<float>(i + 1) + Lstep.cols - 1)))*((*(Ld.ptr<float>(i + 1) + Lstep.cols - 1)) - (*(Ld.ptr<float>(i)+Lstep.cols - 1)));
float yneg = ((*(c.ptr<float>(i - 1) + Lstep.cols - 1)) + (*(c.ptr<float>(i)+Lstep.cols - 1)))*((*(Ld.ptr<float>(i)+Lstep.cols - 1)) - (*(Ld.ptr<float>(i - 1) + Lstep.cols - 1)));
*(Lstep.ptr<float>(i)+Lstep.cols - 1) = 0.5f*stepsize*(-xneg + ypos - yneg);
}
Ld = Ld + Lstep;
}
/* ************************************************************************* */
/**
* @brief This function downsamples the input image with the kernel [1/4,1/2,1/4]
* @param img Input image to be downsampled
* @param dst Output image with half of the resolution of the input image
*/
void downsample_image(const cv::Mat& src, cv::Mat& dst) {
int i1 = 0, j1 = 0, i2 = 0, j2 = 0;
for (i1 = 1; i1 < src.rows; i1 += 2) {
j2 = 0;
for (j1 = 1; j1 < src.cols; j1 += 2) {
*(dst.ptr<float>(i2)+j2) = 0.5f*(*(src.ptr<float>(i1)+j1)) + 0.25f*(*(src.ptr<float>(i1)+j1 - 1) + *(src.ptr<float>(i1)+j1 + 1));
j2++;
}
i2++;
}
}
/* ************************************************************************* */
/**
* @brief This function downsamples the input image using OpenCV resize
* @param img Input image to be downsampled
* @param dst Output image with half of the resolution of the input image
*/
void halfsample_image(const cv::Mat& src, cv::Mat& dst) {
// Make sure the destination image is of the right size
CV_Assert(src.cols / 2 == dst.cols);
CV_Assert(src.rows / 2 == dst.rows);
resize(src, dst, dst.size(), 0, 0, cv::INTER_AREA);
}
/* ************************************************************************* */
/**
* @brief Compute Scharr derivative kernels for sizes different than 3
* @param kx_ The derivative kernel in x-direction
* @param ky_ The derivative kernel in y-direction
* @param dx The derivative order in x-direction
* @param dy The derivative order in y-direction
* @param scale The kernel size
*/
void compute_derivative_kernels(cv::OutputArray kx_, cv::OutputArray ky_, int dx, int dy, int scale) {
const int ksize = 3 + 2 * (scale - 1);
// The usual Scharr kernel
if (scale == 1) {
getDerivKernels(kx_, ky_, dx, dy, 0, true, CV_32F);
return;
}
kx_.create(ksize, 1, CV_32F, -1, true);
ky_.create(ksize, 1, CV_32F, -1, true);
Mat kx = kx_.getMat();
Mat ky = ky_.getMat();
float w = 10.0f / 3.0f;
float norm = 1.0f / (2.0f*scale*(w + 2.0f));
for (int k = 0; k < 2; k++) {
Mat* kernel = k == 0 ? &kx : &ky;
int order = k == 0 ? dx : dy;
float kerI[1000];
for (int t = 0; t < ksize; t++) {
kerI[t] = 0;
}
if (order == 0) {
kerI[0] = norm;
kerI[ksize / 2] = w*norm;
kerI[ksize - 1] = norm;
}
else if (order == 1) {
kerI[0] = -1;
kerI[ksize / 2] = 0;
kerI[ksize - 1] = 1;
}
Mat temp(kernel->rows, kernel->cols, CV_32F, &kerI[0]);
temp.copyTo(*kernel);
}
}

44
modules/features2d/src/akaze/nldiffusion_functions.h
View File

@ -5,7 +5,8 @@
* @author Pablo F. Alcantarilla, Jesus Nuevo * @author Pablo F. Alcantarilla, Jesus Nuevo
*/ */
#pragma once #ifndef AKAZE_NLDIFFUSION_FUNCTIONS_H
#define AKAZE_NLDIFFUSION_FUNCTIONS_H
/* ************************************************************************* */ /* ************************************************************************* */
// Includes // Includes
@ -13,20 +14,27 @@
/* ************************************************************************* */ /* ************************************************************************* */
// Declaration of functions // Declaration of functions
void gaussian_2D_convolution(const cv::Mat& src, cv::Mat& dst, const size_t& ksize_x,
const size_t& ksize_y, const float& sigma); namespace cv {
void image_derivatives_scharr(const cv::Mat& src, cv::Mat& dst, namespace details {
const size_t& xorder, const size_t& yorder); namespace akaze {
void pm_g1(const cv::Mat& Lx, const cv::Mat& Ly, cv::Mat& dst, const float& k);
void pm_g2(const cv::Mat& Lx, const cv::Mat& Ly, cv::Mat& dst, const float& k); void gaussian_2D_convolution(const cv::Mat& src, cv::Mat& dst, int ksize_x, int ksize_y, float sigma);
void weickert_diffusivity(const cv::Mat& Lx, const cv::Mat& Ly, cv::Mat& dst, const float& k); void image_derivatives_scharr(const cv::Mat& src, cv::Mat& dst, int xorder, int yorder);
void charbonnier_diffusivity(const cv::Mat& Lx, const cv::Mat& Ly, cv::Mat& dst, const float& k); void pm_g1(const cv::Mat& Lx, const cv::Mat& Ly, cv::Mat& dst, const float& k);
float compute_k_percentile(const cv::Mat& img, float perc, float gscale, void pm_g2(const cv::Mat& Lx, const cv::Mat& Ly, cv::Mat& dst, const float& k);
size_t nbins, size_t ksize_x, size_t ksize_y); void weickert_diffusivity(const cv::Mat& Lx, const cv::Mat& Ly, cv::Mat& dst, const float& k);
void compute_scharr_derivatives(const cv::Mat& src, cv::Mat& dst, int xorder, int, int scale); void charbonnier_diffusivity(const cv::Mat& Lx, const cv::Mat& Ly, cv::Mat& dst, const float& k);
void nld_step_scalar(cv::Mat& Ld, const cv::Mat& c, cv::Mat& Lstep, const float& stepsize); float compute_k_percentile(const cv::Mat& img, float perc, float gscale, int nbins, int ksize_x, int ksize_y);
void downsample_image(const cv::Mat& src, cv::Mat& dst); void compute_scharr_derivatives(const cv::Mat& src, cv::Mat& dst, int xorder, int, int scale);
void halfsample_image(const cv::Mat& src, cv::Mat& dst); void nld_step_scalar(cv::Mat& Ld, const cv::Mat& c, cv::Mat& Lstep, const float& stepsize);
void compute_derivative_kernels(cv::OutputArray kx_, cv::OutputArray ky_, int dx, int dy, int scale); void downsample_image(const cv::Mat& src, cv::Mat& dst);
bool check_maximum_neighbourhood(const cv::Mat& img, int dsize, float value, void halfsample_image(const cv::Mat& src, cv::Mat& dst);
int row, int col, bool same_img); void compute_derivative_kernels(cv::OutputArray kx_, cv::OutputArray ky_, int dx, int dy, int scale);
bool check_maximum_neighbourhood(const cv::Mat& img, int dsize, float value, int row, int col, bool same_img);
}
}
}
#endif

1
modules/features2d/src/kaze/KAZEFeatures.cpp
View File

@ -26,6 +26,7 @@
// Namespaces // Namespaces
using namespace std; using namespace std;
using namespace cv; using namespace cv;
using namespace cv::details::kaze;
//******************************************************************************* //*******************************************************************************
//******************************************************************************* //*******************************************************************************

584
modules/features2d/src/kaze/nldiffusion_functions.cpp
View File

@ -28,349 +28,355 @@
// Namespaces // Namespaces
using namespace std; using namespace std;
using namespace cv; using namespace cv;
using namespace cv::details::kaze;
//************************************************************************************* //*************************************************************************************
//************************************************************************************* //*************************************************************************************
/** namespace cv {
* @brief This function smoothes an image with a Gaussian kernel namespace details {
* @param src Input image namespace kaze {
* @param dst Output image /**
* @param ksize_x Kernel size in X-direction (horizontal) * @brief This function smoothes an image with a Gaussian kernel
* @param ksize_y Kernel size in Y-direction (vertical) * @param src Input image
* @param sigma Kernel standard deviation * @param dst Output image
*/ * @param ksize_x Kernel size in X-direction (horizontal)
void gaussian_2D_convolution(const cv::Mat& src, cv::Mat& dst, * @param ksize_y Kernel size in Y-direction (vertical)
int ksize_x, int ksize_y, float sigma) { * @param sigma Kernel standard deviation
*/
void gaussian_2D_convolution(const cv::Mat& src, cv::Mat& dst,
int ksize_x, int ksize_y, float sigma) {
int ksize_x_ = 0, ksize_y_ = 0; int ksize_x_ = 0, ksize_y_ = 0;
// Compute an appropriate kernel size according to the specified sigma // Compute an appropriate kernel size according to the specified sigma
if (sigma > ksize_x || sigma > ksize_y || ksize_x == 0 || ksize_y == 0) { if (sigma > ksize_x || sigma > ksize_y || ksize_x == 0 || ksize_y == 0) {
ksize_x_ = (int)ceil(2.0f*(1.0f + (sigma-0.8f)/(0.3f))); ksize_x_ = (int)ceil(2.0f*(1.0f + (sigma - 0.8f) / (0.3f)));
ksize_y_ = ksize_x_; ksize_y_ = ksize_x_;
} }
// The kernel size must be and odd number // The kernel size must be and odd number
if ((ksize_x_ % 2) == 0) { if ((ksize_x_ % 2) == 0) {
ksize_x_ += 1; ksize_x_ += 1;
} }
if ((ksize_y_ % 2) == 0) { if ((ksize_y_ % 2) == 0) {
ksize_y_ += 1; ksize_y_ += 1;
} }
// Perform the Gaussian Smoothing with border replication // Perform the Gaussian Smoothing with border replication
GaussianBlur(src,dst,Size(ksize_x_,ksize_y_),sigma,sigma,cv::BORDER_REPLICATE); GaussianBlur(src, dst, Size(ksize_x_, ksize_y_), sigma, sigma, cv::BORDER_REPLICATE);
} }
//************************************************************************************* //*************************************************************************************
//************************************************************************************* //*************************************************************************************
/** /**
* @brief This function computes the Perona and Malik conductivity coefficient g1 * @brief This function computes the Perona and Malik conductivity coefficient g1
* g1 = exp(-|dL|^2/k^2) * g1 = exp(-|dL|^2/k^2)
* @param Lx First order image derivative in X-direction (horizontal) * @param Lx First order image derivative in X-direction (horizontal)
* @param Ly First order image derivative in Y-direction (vertical) * @param Ly First order image derivative in Y-direction (vertical)
* @param dst Output image * @param dst Output image
* @param k Contrast factor parameter * @param k Contrast factor parameter
*/ */
void pm_g1(const cv::Mat& Lx, const cv::Mat& Ly, cv::Mat& dst, float k) { void pm_g1(const cv::Mat& Lx, const cv::Mat& Ly, cv::Mat& dst, float k) {
cv::exp(-(Lx.mul(Lx) + Ly.mul(Ly))/(k*k),dst); cv::exp(-(Lx.mul(Lx) + Ly.mul(Ly)) / (k*k), dst);
} }
//************************************************************************************* //*************************************************************************************
//************************************************************************************* //*************************************************************************************
/** /**
* @brief This function computes the Perona and Malik conductivity coefficient g2 * @brief This function computes the Perona and Malik conductivity coefficient g2
* g2 = 1 / (1 + dL^2 / k^2) * g2 = 1 / (1 + dL^2 / k^2)
* @param Lx First order image derivative in X-direction (horizontal) * @param Lx First order image derivative in X-direction (horizontal)
* @param Ly First order image derivative in Y-direction (vertical) * @param Ly First order image derivative in Y-direction (vertical)
* @param dst Output image * @param dst Output image
* @param k Contrast factor parameter * @param k Contrast factor parameter
*/ */
void pm_g2(const cv::Mat &Lx, const cv::Mat& Ly, cv::Mat& dst, float k) { void pm_g2(const cv::Mat &Lx, const cv::Mat& Ly, cv::Mat& dst, float k) {
dst = 1./(1. + (Lx.mul(Lx) + Ly.mul(Ly))/(k*k)); dst = 1. / (1. + (Lx.mul(Lx) + Ly.mul(Ly)) / (k*k));
} }
//************************************************************************************* //*************************************************************************************
//************************************************************************************* //*************************************************************************************
/** /**
* @brief This function computes Weickert conductivity coefficient g3 * @brief This function computes Weickert conductivity coefficient g3
* @param Lx First order image derivative in X-direction (horizontal) * @param Lx First order image derivative in X-direction (horizontal)
* @param Ly First order image derivative in Y-direction (vertical) * @param Ly First order image derivative in Y-direction (vertical)
* @param dst Output image * @param dst Output image
* @param k Contrast factor parameter * @param k Contrast factor parameter
* @note For more information check the following paper: J. Weickert * @note For more information check the following paper: J. Weickert
* Applications of nonlinear diffusion in image processing and computer vision, * Applications of nonlinear diffusion in image processing and computer vision,
* Proceedings of Algorithmy 2000 * Proceedings of Algorithmy 2000
*/ */
void weickert_diffusivity(const cv::Mat& Lx, const cv::Mat& Ly, cv::Mat& dst, float k) { void weickert_diffusivity(const cv::Mat& Lx, const cv::Mat& Ly, cv::Mat& dst, float k) {
Mat modg; Mat modg;
cv::pow((Lx.mul(Lx) + Ly.mul(Ly))/(k*k),4,modg); cv::pow((Lx.mul(Lx) + Ly.mul(Ly)) / (k*k), 4, modg);
cv::exp(-3.315/modg, dst); cv::exp(-3.315 / modg, dst);
dst = 1.0f - dst; dst = 1.0f - dst;
} }
//************************************************************************************* //*************************************************************************************
//************************************************************************************* //*************************************************************************************
/** /**
* @brief This function computes a good empirical value for the k contrast factor * @brief This function computes a good empirical value for the k contrast factor
* given an input image, the percentile (0-1), the gradient scale and the number of * given an input image, the percentile (0-1), the gradient scale and the number of
* bins in the histogram * bins in the histogram
* @param img Input image * @param img Input image
* @param perc Percentile of the image gradient histogram (0-1) * @param perc Percentile of the image gradient histogram (0-1)
* @param gscale Scale for computing the image gradient histogram * @param gscale Scale for computing the image gradient histogram
* @param nbins Number of histogram bins * @param nbins Number of histogram bins
* @param ksize_x Kernel size in X-direction (horizontal) for the Gaussian smoothing kernel * @param ksize_x Kernel size in X-direction (horizontal) for the Gaussian smoothing kernel
* @param ksize_y Kernel size in Y-direction (vertical) for the Gaussian smoothing kernel * @param ksize_y Kernel size in Y-direction (vertical) for the Gaussian smoothing kernel
* @return k contrast factor * @return k contrast factor
*/ */
float compute_k_percentile(const cv::Mat& img, float perc, float gscale, float compute_k_percentile(const cv::Mat& img, float perc, float gscale, int nbins, int ksize_x, int ksize_y) {
int nbins, int ksize_x, int ksize_y) {
int nbin = 0, nelements = 0, nthreshold = 0, k = 0; int nbin = 0, nelements = 0, nthreshold = 0, k = 0;
float kperc = 0.0, modg = 0.0, lx = 0.0, ly = 0.0; float kperc = 0.0, modg = 0.0, lx = 0.0, ly = 0.0;
float npoints = 0.0; float npoints = 0.0;
float hmax = 0.0; float hmax = 0.0;
// Create the array for the histogram // Create the array for the histogram
std::vector<int> hist(nbins, 0); std::vector<int> hist(nbins, 0);
// Create the matrices // Create the matrices
Mat gaussian = Mat::zeros(img.rows,img.cols,CV_32F); Mat gaussian = Mat::zeros(img.rows, img.cols, CV_32F);
Mat Lx = Mat::zeros(img.rows,img.cols,CV_32F); Mat Lx = Mat::zeros(img.rows, img.cols, CV_32F);
Mat Ly = Mat::zeros(img.rows,img.cols,CV_32F); Mat Ly = Mat::zeros(img.rows, img.cols, CV_32F);
// Perform the Gaussian convolution // Perform the Gaussian convolution
gaussian_2D_convolution(img,gaussian,ksize_x,ksize_y,gscale); gaussian_2D_convolution(img, gaussian, ksize_x, ksize_y, gscale);
// Compute the Gaussian derivatives Lx and Ly // Compute the Gaussian derivatives Lx and Ly
Scharr(gaussian,Lx,CV_32F,1,0,1,0,cv::BORDER_DEFAULT); Scharr(gaussian, Lx, CV_32F, 1, 0, 1, 0, cv::BORDER_DEFAULT);
Scharr(gaussian,Ly,CV_32F,0,1,1,0,cv::BORDER_DEFAULT); Scharr(gaussian, Ly, CV_32F, 0, 1, 1, 0, cv::BORDER_DEFAULT);
// Skip the borders for computing the histogram // Skip the borders for computing the histogram
for (int i = 1; i < gaussian.rows-1; i++) { for (int i = 1; i < gaussian.rows - 1; i++) {
for (int j = 1; j < gaussian.cols-1; j++) { for (int j = 1; j < gaussian.cols - 1; j++) {
lx = *(Lx.ptr<float>(i)+j); lx = *(Lx.ptr<float>(i)+j);
ly = *(Ly.ptr<float>(i)+j); ly = *(Ly.ptr<float>(i)+j);
modg = sqrt(lx*lx + ly*ly); modg = sqrt(lx*lx + ly*ly);
// Get the maximum // Get the maximum
if (modg > hmax) { if (modg > hmax) {
hmax = modg; hmax = modg;
} }
} }
} }
// Skip the borders for computing the histogram // Skip the borders for computing the histogram
for (int i = 1; i < gaussian.rows-1; i++) { for (int i = 1; i < gaussian.rows - 1; i++) {
for (int j = 1; j < gaussian.cols-1; j++) { for (int j = 1; j < gaussian.cols - 1; j++) {
lx = *(Lx.ptr<float>(i)+j); lx = *(Lx.ptr<float>(i)+j);
ly = *(Ly.ptr<float>(i)+j); ly = *(Ly.ptr<float>(i)+j);
modg = sqrt(lx*lx + ly*ly); modg = sqrt(lx*lx + ly*ly);
// Find the correspondent bin // Find the correspondent bin
if (modg != 0.0) { if (modg != 0.0) {
nbin = (int)floor(nbins*(modg/hmax)); nbin = (int)floor(nbins*(modg / hmax));
if (nbin == nbins) { if (nbin == nbins) {
nbin--; nbin--;
} }
hist[nbin]++; hist[nbin]++;
npoints++; npoints++;
} }
} }
} }
// Now find the perc of the histogram percentile // Now find the perc of the histogram percentile
nthreshold = (size_t)(npoints*perc); nthreshold = (size_t)(npoints*perc);
for (k = 0; nelements < nthreshold && k < nbins; k++) { for (k = 0; nelements < nthreshold && k < nbins; k++) {
nelements = nelements + hist[k]; nelements = nelements + hist[k];
} }
if (nelements < nthreshold) { if (nelements < nthreshold) {
kperc = 0.03f; kperc = 0.03f;
} }
else { else {
kperc = hmax*((float)(k)/(float)nbins); kperc = hmax*((float)(k) / (float)nbins);
} }
return kperc; return kperc;
} }
//************************************************************************************* //*************************************************************************************
//************************************************************************************* //*************************************************************************************
/** /**
* @brief This function computes Scharr image derivatives * @brief This function computes Scharr image derivatives
* @param src Input image * @param src Input image
* @param dst Output image * @param dst Output image
* @param xorder Derivative order in X-direction (horizontal) * @param xorder Derivative order in X-direction (horizontal)
* @param yorder Derivative order in Y-direction (vertical) * @param yorder Derivative order in Y-direction (vertical)
* @param scale Scale factor or derivative size * @param scale Scale factor or derivative size
*/ */
void compute_scharr_derivatives(const cv::Mat& src, cv::Mat& dst, void compute_scharr_derivatives(const cv::Mat& src, cv::Mat& dst,
int xorder, int yorder, int scale) { int xorder, int yorder, int scale) {
Mat kx, ky; Mat kx, ky;
compute_derivative_kernels(kx,ky,xorder,yorder,scale); compute_derivative_kernels(kx, ky, xorder, yorder, scale);
sepFilter2D(src,dst,CV_32F,kx,ky); sepFilter2D(src, dst, CV_32F, kx, ky);
} }
//************************************************************************************* //*************************************************************************************
//************************************************************************************* //*************************************************************************************
/** /**
* @brief Compute derivative kernels for sizes different than 3 * @brief Compute derivative kernels for sizes different than 3
* @param _kx Horizontal kernel values * @param _kx Horizontal kernel values
* @param _ky Vertical kernel values * @param _ky Vertical kernel values
* @param dx Derivative order in X-direction (horizontal) * @param dx Derivative order in X-direction (horizontal)
* @param dy Derivative order in Y-direction (vertical) * @param dy Derivative order in Y-direction (vertical)
* @param scale_ Scale factor or derivative size * @param scale_ Scale factor or derivative size
*/ */
void compute_derivative_kernels(cv::OutputArray _kx, cv::OutputArray _ky, void compute_derivative_kernels(cv::OutputArray _kx, cv::OutputArray _ky,
int dx, int dy, int scale) { int dx, int dy, int scale) {
int ksize = 3 + 2*(scale-1); int ksize = 3 + 2 * (scale - 1);
// The standard Scharr kernel // The standard Scharr kernel
if (scale == 1) { if (scale == 1) {
getDerivKernels(_kx,_ky,dx,dy,0,true,CV_32F); getDerivKernels(_kx, _ky, dx, dy, 0, true, CV_32F);
return; return;
} }
_kx.create(ksize,1,CV_32F,-1,true); _kx.create(ksize, 1, CV_32F, -1, true);
_ky.create(ksize,1,CV_32F,-1,true); _ky.create(ksize, 1, CV_32F, -1, true);
Mat kx = _kx.getMat(); Mat kx = _kx.getMat();
Mat ky = _ky.getMat(); Mat ky = _ky.getMat();
float w = 10.0f/3.0f; float w = 10.0f / 3.0f;
float norm = 1.0f/(2.0f*scale*(w+2.0f)); float norm = 1.0f / (2.0f*scale*(w + 2.0f));
for (int k = 0; k < 2; k++) { for (int k = 0; k < 2; k++) {
Mat* kernel = k == 0 ? &kx : &ky; Mat* kernel = k == 0 ? &kx : &ky;
int order = k == 0 ? dx : dy; int order = k == 0 ? dx : dy;
std::vector<float> kerI(ksize, 0.0f); std::vector<float> kerI(ksize, 0.0f);
if (order == 0) { if (order == 0) {
kerI[0] = norm, kerI[ksize/2] = w*norm, kerI[ksize-1] = norm; kerI[0] = norm, kerI[ksize / 2] = w*norm, kerI[ksize - 1] = norm;
} }
else if (order == 1) { else if (order == 1) {
kerI[0] = -1, kerI[ksize/2] = 0, kerI[ksize-1] = 1; kerI[0] = -1, kerI[ksize / 2] = 0, kerI[ksize - 1] = 1;
} }
Mat temp(kernel->rows,kernel->cols,CV_32F,&kerI[0]); Mat temp(kernel->rows, kernel->cols, CV_32F, &kerI[0]);
temp.copyTo(*kernel); temp.copyTo(*kernel);
} }
} }
//************************************************************************************* //*************************************************************************************
//************************************************************************************* //*************************************************************************************
/** /**
* @brief This function performs a scalar non-linear diffusion step * @brief This function performs a scalar non-linear diffusion step
* @param Ld2 Output image in the evolution * @param Ld2 Output image in the evolution
* @param c Conductivity image * @param c Conductivity image
* @param Lstep Previous image in the evolution * @param Lstep Previous image in the evolution
* @param stepsize The step size in time units * @param stepsize The step size in time units
* @note Forward Euler Scheme 3x3 stencil * @note Forward Euler Scheme 3x3 stencil
* The function c is a scalar value that depends on the gradient norm * The function c is a scalar value that depends on the gradient norm
* dL_by_ds = d(c dL_by_dx)_by_dx + d(c dL_by_dy)_by_dy * dL_by_ds = d(c dL_by_dx)_by_dx + d(c dL_by_dy)_by_dy
*/ */
void nld_step_scalar(cv::Mat& Ld, const cv::Mat& c, cv::Mat& Lstep, float stepsize) { void nld_step_scalar(cv::Mat& Ld, const cv::Mat& c, cv::Mat& Lstep, float stepsize) {
#ifdef _OPENMP #ifdef _OPENMP
#pragma omp parallel for schedule(dynamic) #pragma omp parallel for schedule(dynamic)
#endif #endif
for (int i = 1; i < Lstep.rows-1; i++) { for (int i = 1; i < Lstep.rows - 1; i++) {
for (int j = 1; j < Lstep.cols-1; j++) { for (int j = 1; j < Lstep.cols - 1; j++) {
float xpos = ((*(c.ptr<float>(i)+j))+(*(c.ptr<float>(i)+j+1)))*((*(Ld.ptr<float>(i)+j+1))-(*(Ld.ptr<float>(i)+j))); float xpos = ((*(c.ptr<float>(i)+j)) + (*(c.ptr<float>(i)+j + 1)))*((*(Ld.ptr<float>(i)+j + 1)) - (*(Ld.ptr<float>(i)+j)));
float xneg = ((*(c.ptr<float>(i)+j-1))+(*(c.ptr<float>(i)+j)))*((*(Ld.ptr<float>(i)+j))-(*(Ld.ptr<float>(i)+j-1))); float xneg = ((*(c.ptr<float>(i)+j - 1)) + (*(c.ptr<float>(i)+j)))*((*(Ld.ptr<float>(i)+j)) - (*(Ld.ptr<float>(i)+j - 1)));
float ypos = ((*(c.ptr<float>(i)+j))+(*(c.ptr<float>(i+1)+j)))*((*(Ld.ptr<float>(i+1)+j))-(*(Ld.ptr<float>(i)+j))); float ypos = ((*(c.ptr<float>(i)+j)) + (*(c.ptr<float>(i + 1) + j)))*((*(Ld.ptr<float>(i + 1) + j)) - (*(Ld.ptr<float>(i)+j)));
float yneg = ((*(c.ptr<float>(i-1)+j))+(*(c.ptr<float>(i)+j)))*((*(Ld.ptr<float>(i)+j))-(*(Ld.ptr<float>(i-1)+j))); float yneg = ((*(c.ptr<float>(i - 1) + j)) + (*(c.ptr<float>(i)+j)))*((*(Ld.ptr<float>(i)+j)) - (*(Ld.ptr<float>(i - 1) + j)));
*(Lstep.ptr<float>(i)+j) = 0.5f*stepsize*(xpos-xneg + ypos-yneg); *(Lstep.ptr<float>(i)+j) = 0.5f*stepsize*(xpos - xneg + ypos - yneg);
} }
} }
for (int j = 1; j < Lstep.cols-1; j++) { for (int j = 1; j < Lstep.cols - 1; j++) {
float xpos = ((*(c.ptr<float>(0)+j))+(*(c.ptr<float>(0)+j+1)))*((*(Ld.ptr<float>(0)+j+1))-(*(Ld.ptr<float>(0)+j))); float xpos = ((*(c.ptr<float>(0) + j)) + (*(c.ptr<float>(0) + j + 1)))*((*(Ld.ptr<float>(0) + j + 1)) - (*(Ld.ptr<float>(0) + j)));
float xneg = ((*(c.ptr<float>(0)+j-1))+(*(c.ptr<float>(0)+j)))*((*(Ld.ptr<float>(0)+j))-(*(Ld.ptr<float>(0)+j-1))); float xneg = ((*(c.ptr<float>(0) + j - 1)) + (*(c.ptr<float>(0) + j)))*((*(Ld.ptr<float>(0) + j)) - (*(Ld.ptr<float>(0) + j - 1)));
float ypos = ((*(c.ptr<float>(0)+j))+(*(c.ptr<float>(1)+j)))*((*(Ld.ptr<float>(1)+j))-(*(Ld.ptr<float>(0)+j))); float ypos = ((*(c.ptr<float>(0) + j)) + (*(c.ptr<float>(1) + j)))*((*(Ld.ptr<float>(1) + j)) - (*(Ld.ptr<float>(0) + j)));
float yneg = ((*(c.ptr<float>(0)+j))+(*(c.ptr<float>(0)+j)))*((*(Ld.ptr<float>(0)+j))-(*(Ld.ptr<float>(0)+j))); float yneg = ((*(c.ptr<float>(0) + j)) + (*(c.ptr<float>(0) + j)))*((*(Ld.ptr<float>(0) + j)) - (*(Ld.ptr<float>(0) + j)));
*(Lstep.ptr<float>(0)+j) = 0.5f*stepsize*(xpos-xneg + ypos-yneg); *(Lstep.ptr<float>(0) + j) = 0.5f*stepsize*(xpos - xneg + ypos - yneg);
} }
for (int j = 1; j < Lstep.cols-1; j++) { for (int j = 1; j < Lstep.cols - 1; j++) {
float xpos = ((*(c.ptr<float>(Lstep.rows-1)+j))+(*(c.ptr<float>(Lstep.rows-1)+j+1)))*((*(Ld.ptr<float>(Lstep.rows-1)+j+1))-(*(Ld.ptr<float>(Lstep.rows-1)+j))); float xpos = ((*(c.ptr<float>(Lstep.rows - 1) + j)) + (*(c.ptr<float>(Lstep.rows - 1) + j + 1)))*((*(Ld.ptr<float>(Lstep.rows - 1) + j + 1)) - (*(Ld.ptr<float>(Lstep.rows - 1) + j)));
float xneg = ((*(c.ptr<float>(Lstep.rows-1)+j-1))+(*(c.ptr<float>(Lstep.rows-1)+j)))*((*(Ld.ptr<float>(Lstep.rows-1)+j))-(*(Ld.ptr<float>(Lstep.rows-1)+j-1))); float xneg = ((*(c.ptr<float>(Lstep.rows - 1) + j - 1)) + (*(c.ptr<float>(Lstep.rows - 1) + j)))*((*(Ld.ptr<float>(Lstep.rows - 1) + j)) - (*(Ld.ptr<float>(Lstep.rows - 1) + j - 1)));
float ypos = ((*(c.ptr<float>(Lstep.rows-1)+j))+(*(c.ptr<float>(Lstep.rows-1)+j)))*((*(Ld.ptr<float>(Lstep.rows-1)+j))-(*(Ld.ptr<float>(Lstep.rows-1)+j))); float ypos = ((*(c.ptr<float>(Lstep.rows - 1) + j)) + (*(c.ptr<float>(Lstep.rows - 1) + j)))*((*(Ld.ptr<float>(Lstep.rows - 1) + j)) - (*(Ld.ptr<float>(Lstep.rows - 1) + j)));
float yneg = ((*(c.ptr<float>(Lstep.rows-2)+j))+(*(c.ptr<float>(Lstep.rows-1)+j)))*((*(Ld.ptr<float>(Lstep.rows-1)+j))-(*(Ld.ptr<float>(Lstep.rows-2)+j))); float yneg = ((*(c.ptr<float>(Lstep.rows - 2) + j)) + (*(c.ptr<float>(Lstep.rows - 1) + j)))*((*(Ld.ptr<float>(Lstep.rows - 1) + j)) - (*(Ld.ptr<float>(Lstep.rows - 2) + j)));
*(Lstep.ptr<float>(Lstep.rows-1)+j) = 0.5f*stepsize*(xpos-xneg + ypos-yneg); *(Lstep.ptr<float>(Lstep.rows - 1) + j) = 0.5f*stepsize*(xpos - xneg + ypos - yneg);
} }
for (int i = 1; i < Lstep.rows-1; i++) { for (int i = 1; i < Lstep.rows - 1; i++) {
float xpos = ((*(c.ptr<float>(i)))+(*(c.ptr<float>(i)+1)))*((*(Ld.ptr<float>(i)+1))-(*(Ld.ptr<float>(i)))); float xpos = ((*(c.ptr<float>(i))) + (*(c.ptr<float>(i)+1)))*((*(Ld.ptr<float>(i)+1)) - (*(Ld.ptr<float>(i))));
float xneg = ((*(c.ptr<float>(i)))+(*(c.ptr<float>(i))))*((*(Ld.ptr<float>(i)))-(*(Ld.ptr<float>(i)))); float xneg = ((*(c.ptr<float>(i))) + (*(c.ptr<float>(i))))*((*(Ld.ptr<float>(i))) - (*(Ld.ptr<float>(i))));
float ypos = ((*(c.ptr<float>(i)))+(*(c.ptr<float>(i+1))))*((*(Ld.ptr<float>(i+1)))-(*(Ld.ptr<float>(i)))); float ypos = ((*(c.ptr<float>(i))) + (*(c.ptr<float>(i + 1))))*((*(Ld.ptr<float>(i + 1))) - (*(Ld.ptr<float>(i))));
float yneg = ((*(c.ptr<float>(i-1)))+(*(c.ptr<float>(i))))*((*(Ld.ptr<float>(i)))-(*(Ld.ptr<float>(i-1)))); float yneg = ((*(c.ptr<float>(i - 1))) + (*(c.ptr<float>(i))))*((*(Ld.ptr<float>(i))) - (*(Ld.ptr<float>(i - 1))));
*(Lstep.ptr<float>(i)) = 0.5f*stepsize*(xpos-xneg + ypos-yneg); *(Lstep.ptr<float>(i)) = 0.5f*stepsize*(xpos - xneg + ypos - yneg);
} }
for (int i = 1; i < Lstep.rows-1; i++) { for (int i = 1; i < Lstep.rows - 1; i++) {
float xpos = ((*(c.ptr<float>(i)+Lstep.cols-1))+(*(c.ptr<float>(i)+Lstep.cols-1)))*((*(Ld.ptr<float>(i)+Lstep.cols-1))-(*(Ld.ptr<float>(i)+Lstep.cols-1))); float xpos = ((*(c.ptr<float>(i)+Lstep.cols - 1)) + (*(c.ptr<float>(i)+Lstep.cols - 1)))*((*(Ld.ptr<float>(i)+Lstep.cols - 1)) - (*(Ld.ptr<float>(i)+Lstep.cols - 1)));
float xneg = ((*(c.ptr<float>(i)+Lstep.cols-2))+(*(c.ptr<float>(i)+Lstep.cols-1)))*((*(Ld.ptr<float>(i)+Lstep.cols-1))-(*(Ld.ptr<float>(i)+Lstep.cols-2))); float xneg = ((*(c.ptr<float>(i)+Lstep.cols - 2)) + (*(c.ptr<float>(i)+Lstep.cols - 1)))*((*(Ld.ptr<float>(i)+Lstep.cols - 1)) - (*(Ld.ptr<float>(i)+Lstep.cols - 2)));
float ypos = ((*(c.ptr<float>(i)+Lstep.cols-1))+(*(c.ptr<float>(i+1)+Lstep.cols-1)))*((*(Ld.ptr<float>(i+1)+Lstep.cols-1))-(*(Ld.ptr<float>(i)+Lstep.cols-1))); float ypos = ((*(c.ptr<float>(i)+Lstep.cols - 1)) + (*(c.ptr<float>(i + 1) + Lstep.cols - 1)))*((*(Ld.ptr<float>(i + 1) + Lstep.cols - 1)) - (*(Ld.ptr<float>(i)+Lstep.cols - 1)));
float yneg = ((*(c.ptr<float>(i-1)+Lstep.cols-1))+(*(c.ptr<float>(i)+Lstep.cols-1)))*((*(Ld.ptr<float>(i)+Lstep.cols-1))-(*(Ld.ptr<float>(i-1)+Lstep.cols-1))); float yneg = ((*(c.ptr<float>(i - 1) + Lstep.cols - 1)) + (*(c.ptr<float>(i)+Lstep.cols - 1)))*((*(Ld.ptr<float>(i)+Lstep.cols - 1)) - (*(Ld.ptr<float>(i - 1) + Lstep.cols - 1)));
*(Lstep.ptr<float>(i)+Lstep.cols-1) = 0.5f*stepsize*(xpos-xneg + ypos-yneg); *(Lstep.ptr<float>(i)+Lstep.cols - 1) = 0.5f*stepsize*(xpos - xneg + ypos - yneg);
} }
Ld = Ld + Lstep; Ld = Ld + Lstep;
}
//*************************************************************************************
//*************************************************************************************
/**
* @brief This function checks if a given pixel is a maximum in a local neighbourhood
* @param img Input image where we will perform the maximum search
* @param dsize Half size of the neighbourhood
* @param value Response value at (x,y) position
* @param row Image row coordinate
* @param col Image column coordinate
* @param same_img Flag to indicate if the image value at (x,y) is in the input image
* @return 1->is maximum, 0->otherwise
*/
bool check_maximum_neighbourhood(const cv::Mat& img, int dsize, float value,
int row, int col, bool same_img) {
bool response = true;
for (int i = row-dsize; i <= row+dsize; i++) {
for (int j = col-dsize; j <= col+dsize; j++) {
if (i >= 0 && i < img.rows && j >= 0 && j < img.cols) {
if (same_img == true) {
if (i != row || j != col) {
if ((*(img.ptr<float>(i)+j)) > value) {
response = false;
return response;
} }
}
}
else {
if ((*(img.ptr<float>(i)+j)) > value) {
response = false;
return response;
}
}
}
}
}
return response; //*************************************************************************************
} //*************************************************************************************
/**
* @brief This function checks if a given pixel is a maximum in a local neighbourhood
* @param img Input image where we will perform the maximum search
* @param dsize Half size of the neighbourhood
* @param value Response value at (x,y) position
* @param row Image row coordinate
* @param col Image column coordinate
* @param same_img Flag to indicate if the image value at (x,y) is in the input image
* @return 1->is maximum, 0->otherwise
*/
bool check_maximum_neighbourhood(const cv::Mat& img, int dsize, float value,
int row, int col, bool same_img) {
bool response = true;
for (int i = row - dsize; i <= row + dsize; i++) {
for (int j = col - dsize; j <= col + dsize; j++) {
if (i >= 0 && i < img.rows && j >= 0 && j < img.cols) {
if (same_img == true) {
if (i != row || j != col) {
if ((*(img.ptr<float>(i)+j)) > value) {
response = false;
return response;
}
}
}
else {
if ((*(img.ptr<float>(i)+j)) > value) {
response = false;
return response;
}
}
}
}
}
return response;
}
}
}
}

54
modules/features2d/src/kaze/nldiffusion_functions.h
View File

@ -1,4 +1,3 @@
/** /**
* @file nldiffusion_functions.h * @file nldiffusion_functions.h
* @brief Functions for non-linear diffusion applications: * @brief Functions for non-linear diffusion applications:
@ -9,43 +8,40 @@
* @author Pablo F. Alcantarilla * @author Pablo F. Alcantarilla
*/ */
#ifndef NLDIFFUSION_FUNCTIONS_H_ #ifndef KAZE_NLDIFFUSION_FUNCTIONS_H
#define NLDIFFUSION_FUNCTIONS_H_ #define KAZE_NLDIFFUSION_FUNCTIONS_H
//******************************************************************************
//******************************************************************************
// Includes // Includes
#include "config.h" #include "precomp.hpp"
//************************************************************************************* //*************************************************************************************
//************************************************************************************* //*************************************************************************************
// Gaussian 2D convolution namespace cv {
void gaussian_2D_convolution(const cv::Mat& src, cv::Mat& dst, namespace details {
int ksize_x, int ksize_y, float sigma); namespace kaze {
// Diffusivity functions // Gaussian 2D convolution
void pm_g1(const cv::Mat& Lx, const cv::Mat& Ly, cv::Mat& dst, float k); void gaussian_2D_convolution(const cv::Mat& src, cv::Mat& dst, int ksize_x, int ksize_y, float sigma);
void pm_g2(const cv::Mat& Lx, const cv::Mat& Ly, cv::Mat& dst, float k);
void weickert_diffusivity(const cv::Mat& Lx, const cv::Mat& Ly, cv::Mat& dst, float k);
float compute_k_percentile(const cv::Mat& img, float perc, float gscale,
int nbins, int ksize_x, int ksize_y);
// Image derivatives // Diffusivity functions
void compute_scharr_derivatives(const cv::Mat& src, cv::Mat& dst, void pm_g1(const cv::Mat& Lx, const cv::Mat& Ly, cv::Mat& dst, float k);
int xorder, int yorder, int scale); void pm_g2(const cv::Mat& Lx, const cv::Mat& Ly, cv::Mat& dst, float k);
void compute_derivative_kernels(cv::OutputArray _kx, cv::OutputArray _ky, void weickert_diffusivity(const cv::Mat& Lx, const cv::Mat& Ly, cv::Mat& dst, float k);
int dx, int dy, int scale); float compute_k_percentile(const cv::Mat& img, float perc, float gscale, int nbins, int ksize_x, int ksize_y);
// Nonlinear diffusion filtering scalar step // Image derivatives
void nld_step_scalar(cv::Mat& Ld, const cv::Mat& c, cv::Mat& Lstep, float stepsize); void compute_scharr_derivatives(const cv::Mat& src, cv::Mat& dst, int xorder, int yorder, int scale);
void compute_derivative_kernels(cv::OutputArray _kx, cv::OutputArray _ky, int dx, int dy, int scale);
// For non-maxima suppresion // Nonlinear diffusion filtering scalar step
bool check_maximum_neighbourhood(const cv::Mat& img, int dsize, float value, void nld_step_scalar(cv::Mat& Ld, const cv::Mat& c, cv::Mat& Lstep, float stepsize);
int row, int col, bool same_img);
//************************************************************************************* // For non-maxima suppresion
//************************************************************************************* bool check_maximum_neighbourhood(const cv::Mat& img, int dsize, float value, int row, int col, bool same_img);
#endif // NLDIFFUSION_FUNCTIONS_H_ }
}
}
#endif