diff --git a/modules/photo/include/opencv2/photo.hpp b/modules/photo/include/opencv2/photo.hpp
index df0ccbeb34..ee6b12e9dc 100644
--- a/modules/photo/include/opencv2/photo.hpp
+++ b/modules/photo/include/opencv2/photo.hpp
@@ -315,7 +315,7 @@ CV_EXPORTS_W void illuminationChange(InputArray src, InputArray mask, OutputArra
         float alpha = 0.2f, float beta = 0.4f);
 
 CV_EXPORTS_W void textureFlattening(InputArray src, InputArray mask, OutputArray dst,
-        double low_threshold = 30, double high_threshold = 45,
+        float low_threshold = 30, float high_threshold = 45,
         int kernel_size = 3);
 
 CV_EXPORTS_W void edgePreservingFilter(InputArray src, OutputArray dst, int flags = 1,
diff --git a/modules/photo/src/seamless_cloning.cpp b/modules/photo/src/seamless_cloning.cpp
index 445c6dae74..2564145e58 100644
--- a/modules/photo/src/seamless_cloning.cpp
+++ b/modules/photo/src/seamless_cloning.cpp
@@ -41,7 +41,6 @@
 
 #include "precomp.hpp"
 #include "opencv2/photo.hpp"
-#include <stdlib.h>
 
 #include "seamless_cloning.hpp"
 
@@ -50,9 +49,9 @@ using namespace cv;
 
 void cv::seamlessClone(InputArray _src, InputArray _dst, InputArray _mask, Point p, OutputArray _blend, int flags)
 {
-    Mat src  = _src.getMat();
-    Mat dest = _dst.getMat();
-    Mat mask = _mask.getMat();
+    const Mat src  = _src.getMat();
+    const Mat dest = _dst.getMat();
+    const Mat mask = _mask.getMat();
     _blend.create(dest.size(), CV_8UC3);
     Mat blend = _blend.getMat();
 
@@ -87,6 +86,8 @@ void cv::seamlessClone(InputArray _src, InputArray _dst, InputArray _mask, Point
     int lenx = maxx - minx;
     int leny = maxy - miny;
 
+    Mat patch = Mat::zeros(Size(leny, lenx), CV_8UC3);
+
     int minxd = p.y - lenx/2;
     int maxxd = p.y + lenx/2;
     int minyd = p.x - leny/2;
@@ -102,12 +103,14 @@ void cv::seamlessClone(InputArray _src, InputArray _dst, InputArray _mask, Point
 
     gray(roi_s).copyTo(destinationROI);
     src(roi_s).copyTo(sourceROI,gray(roi_s));
+    src(roi_s).copyTo(patch, gray(roi_s));
 
     destinationROI = cd_mask(roi_d);
     cs_mask(roi_s).copyTo(destinationROI);
 
+
     Cloning obj;
-    obj.normal_clone(dest,cd_mask,dst_mask,blend,flags);
+    obj.normalClone(dest,cd_mask,dst_mask,blend,flags);
 
 }
 
@@ -134,7 +137,7 @@ void cv::colorChange(InputArray _src, InputArray _mask, OutputArray _dst, float
     src.copyTo(cs_mask,gray);
 
     Cloning obj;
-    obj.local_color_change(src,cs_mask,gray,blend,red,green,blue);
+    obj.localColorChange(src,cs_mask,gray,blend,red,green,blue);
 }
 
 void cv::illuminationChange(InputArray _src, InputArray _mask, OutputArray _dst, float a, float b)
@@ -159,12 +162,12 @@ void cv::illuminationChange(InputArray _src, InputArray _mask, OutputArray _dst,
     src.copyTo(cs_mask,gray);
 
     Cloning obj;
-    obj.illum_change(src,cs_mask,gray,blend,alpha,beta);
+    obj.illuminationChange(src,cs_mask,gray,blend,alpha,beta);
 
 }
 
 void cv::textureFlattening(InputArray _src, InputArray _mask, OutputArray _dst,
-                           double low_threshold, double high_threshold, int kernel_size)
+                           float low_threshold, float high_threshold, int kernel_size)
 {
 
     Mat src  = _src.getMat();
@@ -184,5 +187,5 @@ void cv::textureFlattening(InputArray _src, InputArray _mask, OutputArray _dst,
     src.copyTo(cs_mask,gray);
 
     Cloning obj;
-    obj.texture_flatten(src,cs_mask,gray,low_threshold,high_threshold,kernel_size,blend);
+    obj.textureFlatten(src,cs_mask,gray,low_threshold,high_threshold,kernel_size,blend);
 }
diff --git a/modules/photo/src/seamless_cloning.hpp b/modules/photo/src/seamless_cloning.hpp
index f0713ceb2e..a0f9d6863c 100644
--- a/modules/photo/src/seamless_cloning.hpp
+++ b/modules/photo/src/seamless_cloning.hpp
@@ -39,546 +39,52 @@
 //
 //M*/
 
+#ifndef CV_SEAMLESS_CLONING_HPP___
+#define CV_SEAMLESS_CLONING_HPP___
+
 #include "precomp.hpp"
 #include "opencv2/photo.hpp"
-#include <iostream>
-#include <stdlib.h>
-#include <complex>
-#include "math.h"
 
-using namespace std;
-using namespace cv;
+#include <vector>
 
-class Cloning
+namespace cv
 {
 
-    public:
-
-        vector <Mat> rgb_channel, rgbx_channel, rgby_channel, output;
-        Mat grx, gry, sgx, sgy, srx32, sry32, grx32, gry32, smask, smask1;
-        void init_var(Mat &I, Mat &wmask);
-        void initialization(Mat &I, Mat &mask, Mat &wmask);
-        void scalar_product(Mat mat, float r, float g, float b);
-        void array_product(Mat mat1, Mat mat2, Mat mat3);
-        void poisson(Mat &I, Mat &gx, Mat &gy, Mat &sx, Mat &sy);
-        void evaluate(Mat &I, Mat &wmask, Mat &cloned);
-        void getGradientx(const Mat &img, Mat &gx);
-        void getGradienty(const Mat &img, Mat &gy);
-        void lapx(const Mat &img, Mat &gxx);
-        void lapy(const Mat &img, Mat &gyy);
-        void dst(double *mod_diff, double *sineTransform,int h,int w);
-        void idst(double *mod_diff, double *sineTransform,int h,int w);
-        void transpose(double *mat, double *mat_t,int h,int w);
-        void solve(const Mat &img, double *mod_diff, Mat &result);
-        void poisson_solver(const Mat &img, Mat &gxx , Mat &gyy, Mat &result);
-        void normal_clone(Mat &I, Mat &mask, Mat &wmask, Mat &cloned, int num);
-        void local_color_change(Mat &I, Mat &mask, Mat &wmask, Mat &cloned, float red_mul, float green_mul, float blue_mul);
-        void illum_change(Mat &I, Mat &mask, Mat &wmask, Mat &cloned, float alpha, float beta);
-        void texture_flatten(Mat &I, Mat &mask, Mat &wmask, double low_threshold, double high_threhold, int kernel_size, Mat &cloned);
-};
-
-void Cloning::getGradientx( const Mat &img, Mat &gx)
-{
-    Mat kernel = Mat::zeros(1, 3, CV_8S);
-    kernel.at<char>(0,2) = 1;
-    kernel.at<char>(0,1) = -1;
-    filter2D(img, gx, CV_32F, kernel);
-}
-
-void Cloning::getGradienty( const Mat &img, Mat &gy)
-{
-    Mat kernel = Mat::zeros(3, 1, CV_8S);
-    kernel.at<char>(2,0) = 1;
-    kernel.at<char>(1,0) = -1;
-    filter2D(img, gy, CV_32F, kernel);
-}
-
-void Cloning::lapx( const Mat &img, Mat &gxx)
-{
-    Mat kernel = Mat::zeros(1, 3, CV_8S);
-    kernel.at<char>(0,0) = -1;
-    kernel.at<char>(0,1) = 1;
-    filter2D(img, gxx, CV_32F, kernel);
-}
-
-void Cloning::lapy( const Mat &img, Mat &gyy)
-{
-    Mat kernel = Mat::zeros(3, 1, CV_8S);
-    kernel.at<char>(0,0) = -1;
-    kernel.at<char>(1,0) = 1;
-    filter2D(img, gyy, CV_32F, kernel);
-}
-
-void Cloning::dst(double *mod_diff, double *sineTransform,int h,int w)
-{
-
-    unsigned long int idx;
-
-    Mat temp = Mat(2*h+2,1,CV_32F);
-    Mat res  = Mat(h,1,CV_32F);
-
-    Mat planes[] = {Mat_<float>(temp), Mat::zeros(temp.size(), CV_32F)};
-
-    Mat result;
-    int p=0;
-    for(int i=0;i<w;i++)
+    class Cloning
     {
-        temp.at<float>(0,0) = 0.0;
+        public:
+            void normalClone(const cv::Mat& destination, const cv::Mat &mask, const cv::Mat &wmask, cv::Mat &cloned, int flag);
+            void illuminationChange(cv::Mat &I, cv::Mat &mask, cv::Mat &wmask, cv::Mat &cloned, float alpha, float beta);
+            void localColorChange(cv::Mat &I, cv::Mat &mask, cv::Mat &wmask, cv::Mat &cloned, float red_mul, float green_mul, float blue_mul);
+            void textureFlatten(cv::Mat &I, cv::Mat &mask, cv::Mat &wmask, float low_threshold, float high_threhold, int kernel_size, cv::Mat &cloned);
 
-        for(int j=0,r=1;j<h;j++,r++)
-        {
-            idx = j*w+i;
-            temp.at<float>(r,0) = (float) mod_diff[idx];
-        }
+        protected:
 
-        temp.at<float>(h+1,0)=0.0;
+            void initVariables(const cv::Mat &destination, const cv::Mat &binaryMask);
+            void computeDerivatives(const cv::Mat &destination, const cv::Mat &patch, const cv::Mat &binaryMask);
+            void scalarProduct(cv::Mat mat, float r, float g, float b);
+            void poisson(const cv::Mat &destination);
+            void evaluate(const cv::Mat &I, const cv::Mat &wmask, const cv::Mat &cloned);
+            void dst(const Mat& src, Mat& dest, bool invert = false);
+            void idst(const Mat& src, Mat& dest);
+            void solve(const Mat &img, Mat& mod_diff, Mat &result);
 
-        for(int j=h-1, r=h+2;j>=0;j--,r++)
-        {
-            idx = j*w+i;
-            temp.at<float>(r,0) = (float) (-1.0 * mod_diff[idx]);
-        }
+            void poissonSolver(const cv::Mat &img, cv::Mat &gxx , cv::Mat &gyy, cv::Mat &result);
 
-        merge(planes, 2, result);
+            void arrayProduct(const cv::Mat& lhs, const cv::Mat& rhs, cv::Mat& result) const;
 
-        dft(result,result,0,0);
+            void computeGradientX(const cv::Mat &img, cv::Mat &gx);
+            void computeGradientY(const cv::Mat &img, cv::Mat &gy);
+            void computeLaplacianX(const cv::Mat &img, cv::Mat &gxx);
+            void computeLaplacianY(const cv::Mat &img, cv::Mat &gyy);
 
-        Mat planes1[] = {Mat::zeros(result.size(), CV_32F), Mat::zeros(result.size(), CV_32F)};
+        private:
+            std::vector <cv::Mat> rgbx_channel, rgby_channel, output;
+            cv::Mat destinationGradientX, destinationGradientY;
+            cv::Mat patchGradientX, patchGradientY;
+            cv::Mat binaryMaskFloat, binaryMaskFloatInverted;
 
-        split(result, planes1);
-
-        std::complex<double> two_i = std::sqrt(std::complex<double>(-1));
-
-        double factor = -2*imag(two_i);
-
-        for(int c=1,z=0;c<h+1;c++,z++)
-        {
-            res.at<float>(z,0) = (float) (planes1[1].at<float>(c,0)/factor);
-        }
-
-        for(int q=0,z=0;q<h;q++,z++)
-        {
-            idx = q*w+p;
-            sineTransform[idx] =  res.at<float>(z,0);
-        }
-        p++;
-    }
-}
-
-void Cloning::idst(double *mod_diff, double *sineTransform,int h,int w)
-{
-    int nn = h+1;
-    unsigned long int idx;
-    dst(mod_diff,sineTransform,h,w);
-    for(int  i= 0;i<h;i++)
-        for(int j=0;j<w;j++)
-        {
-            idx = i*w + j;
-            sineTransform[idx] = (double) (2*sineTransform[idx])/nn;
-        }
-
-}
-
-void Cloning::transpose(double *mat, double *mat_t,int h,int w)
-{
-
-    Mat tmp = Mat(h,w,CV_32FC1);
-    unsigned long int idx;
-    for(int i = 0 ; i < h;i++)
-    {
-        for(int j = 0 ; j < w; j++)
-        {
-
-            idx = i*(w) + j;
-            tmp.at<float>(i,j) = (float) mat[idx];
-        }
-    }
-    Mat tmp_t = tmp.t();
-
-    for(int i = 0;i < tmp_t.size().height; i++)
-        for(int j=0;j<tmp_t.size().width;j++)
-        {
-            idx = i*tmp_t.size().width + j;
-            mat_t[idx] = tmp_t.at<float>(i,j);
-        }
-}
-
-void Cloning::solve(const Mat &img, double *mod_diff, Mat &result)
-{
-    int w = img.size().width;
-    int h = img.size().height;
-
-    unsigned long int idx,idx1;
-
-    double *sineTransform = new double[(h-2)*(w-2)];
-    double *sineTransform_t = new double[(h-2)*(w-2)];
-    double *denom = new double[(h-2)*(w-2)];
-    double *invsineTransform = new double[(h-2)*(w-2)];
-    double *invsineTransform_t = new double[(h-2)*(w-2)];
-    double *img_d = new double[(h)*(w)];
-
-    dst(mod_diff,sineTransform,h-2,w-2);
-
-    transpose(sineTransform,sineTransform_t,h-2,w-2);
-
-    dst(sineTransform_t,sineTransform,w-2,h-2);
-
-    transpose(sineTransform,sineTransform_t,w-2,h-2);
-
-    int cy = 1;
-
-    for(int i = 0 ; i < w-2;i++,cy++)
-    {
-        for(int j = 0,cx = 1; j < h-2; j++,cx++)
-        {
-            idx = j*(w-2) + i;
-            denom[idx] = (float) 2*cos(CV_PI*cy/( (double) (w-1))) - 2 + 2*cos(CV_PI*cx/((double) (h-1))) - 2;
-
-        }
-    }
-
-    for(idx = 0 ; idx < (unsigned)(w-2)*(h-2) ;idx++)
-    {
-        sineTransform_t[idx] = sineTransform_t[idx]/denom[idx];
-    }
-
-    idst(sineTransform_t,invsineTransform,h-2,w-2);
-
-    transpose(invsineTransform,invsineTransform_t,h-2,w-2);
-
-    idst(invsineTransform_t,invsineTransform,w-2,h-2);
-
-    transpose(invsineTransform,invsineTransform_t,w-2,h-2);
-
-    for(int i = 0 ; i < h;i++)
-    {
-        for(int j = 0 ; j < w; j++)
-        {
-            idx = i*w + j;
-            img_d[idx] = (double)img.at<uchar>(i,j);
-        }
-    }
-    for(int i = 1 ; i < h-1;i++)
-    {
-        for(int j = 1 ; j < w-1; j++)
-        {
-            idx = i*w + j;
-            img_d[idx] = 0.0;
-        }
-    }
-    for(int i = 1,id1=0 ; i < h-1;i++,id1++)
-    {
-        for(int j = 1,id2=0 ; j < w-1; j++,id2++)
-        {
-            idx = i*w + j;
-            idx1= id1*(w-2) + id2;
-            img_d[idx] = invsineTransform_t[idx1];
-        }
-    }
-
-    for(int i = 0 ; i < h;i++)
-    {
-        for(int j = 0 ; j < w; j++)
-        {
-            idx = i*w + j;
-            if(img_d[idx] < 0.0)
-                result.at<uchar>(i,j) = 0;
-            else if(img_d[idx] > 255.0)
-                result.at<uchar>(i,j) = 255;
-            else
-                result.at<uchar>(i,j) = (uchar) img_d[idx];
-        }
-    }
-
-    delete [] sineTransform;
-    delete [] sineTransform_t;
-    delete [] denom;
-    delete [] invsineTransform;
-    delete [] invsineTransform_t;
-    delete [] img_d;
-}
-
-void Cloning::poisson_solver(const Mat &img, Mat &gxx , Mat &gyy, Mat &result)
-{
-
-    int w = img.size().width;
-    int h = img.size().height;
-
-    unsigned long int idx;
-
-    Mat lap = Mat(img.size(),CV_32FC1);
-
-    lap = gxx + gyy;
-
-    Mat bound = img.clone();
-
-    rectangle(bound, Point(1, 1), Point(img.cols-2, img.rows-2), Scalar::all(0), -1);
-
-    double *boundary_point = new double[h*w];
-
-    for(int i =1;i<h-1;i++)
-        for(int j=1;j<w-1;j++)
-        {
-            idx=i*w + j;
-            boundary_point[idx] = -4*(int)bound.at<uchar>(i,j) + (int)bound.at<uchar>(i,(j+1)) + (int)bound.at<uchar>(i,(j-1))
-                + (int)bound.at<uchar>(i-1,j) + (int)bound.at<uchar>(i+1,j);
-        }
-
-    Mat diff = Mat(h,w,CV_32FC1);
-    for(int i =0;i<h;i++)
-    {
-        for(int j=0;j<w;j++)
-        {
-            idx = i*w+j;
-            diff.at<float>(i,j) = (float) (lap.at<float>(i,j) - boundary_point[idx]);
-        }
-    }
-
-    double *mod_diff = new double[(h-2)*(w-2)];
-    for(int i = 0 ; i < h-2;i++)
-    {
-        for(int j = 0 ; j < w-2; j++)
-        {
-            idx = i*(w-2) + j;
-            mod_diff[idx] = diff.at<float>(i+1,j+1);
-
-        }
-    }
-    ///////////////////////////////////////////////////// Find DST  /////////////////////////////////////////////////////
-
-    solve(img,mod_diff,result);
-
-    delete [] mod_diff;
-    delete [] boundary_point;
-}
-
-void Cloning::init_var(Mat &I, Mat &wmask)
-{
-    grx = Mat(I.size(),CV_32FC3);
-    gry = Mat(I.size(),CV_32FC3);
-    sgx = Mat(I.size(),CV_32FC3);
-    sgy = Mat(I.size(),CV_32FC3);
-
-    split(I,rgb_channel);
-
-    smask = Mat(wmask.size(),CV_32FC1);
-    srx32 = Mat(I.size(),CV_32FC3);
-    sry32 = Mat(I.size(),CV_32FC3);
-    smask1 = Mat(wmask.size(),CV_32FC1);
-    grx32 = Mat(I.size(),CV_32FC3);
-    gry32 = Mat(I.size(),CV_32FC3);
-}
-
-void Cloning::initialization(Mat &I, Mat &mask, Mat &wmask)
-{
-    init_var(I,wmask);
-
-    getGradientx(I,grx);
-    getGradienty(I,gry);
-
-    getGradientx(mask,sgx);
-    getGradienty(mask,sgy);
-
-    Mat Kernel(Size(3, 3), CV_8UC1);
-    Kernel.setTo(Scalar(1));
-
-    erode(wmask, wmask, Kernel, Point(-1,-1), 3);
-
-    wmask.convertTo(smask,CV_32FC1,1.0/255.0);
-    I.convertTo(srx32,CV_32FC3,1.0/255.0);
-    I.convertTo(sry32,CV_32FC3,1.0/255.0);
-}
-
-void Cloning::scalar_product(Mat mat, float r, float g, float b)
-{
-    vector <Mat> channels;
-    split(mat,channels);
-    multiply(channels[2],r,channels[2]);
-    multiply(channels[1],g,channels[1]);
-    multiply(channels[0],b,channels[0]);
-    merge(channels,mat);
-}
-
-void Cloning::array_product(Mat mat1, Mat mat2, Mat mat3)
-{
-    vector <Mat> channels_temp1;
-    vector <Mat> channels_temp2;
-    split(mat1,channels_temp1);
-    split(mat2,channels_temp2);
-    multiply(channels_temp2[2],mat3,channels_temp1[2]);
-    multiply(channels_temp2[1],mat3,channels_temp1[1]);
-    multiply(channels_temp2[0],mat3,channels_temp1[0]);
-    merge(channels_temp1,mat1);
-}
-
-void Cloning::poisson(Mat &I, Mat &gx, Mat &gy, Mat &sx, Mat &sy)
-{
-    Mat fx = Mat(I.size(),CV_32FC3);
-    Mat fy = Mat(I.size(),CV_32FC3);
-
-    fx = gx + sx;
-    fy = gy + sy;
-
-    Mat gxx = Mat(I.size(),CV_32FC3);
-    Mat gyy = Mat(I.size(),CV_32FC3);
-
-    lapx(fx,gxx);
-    lapy(fy,gyy);
-
-    split(gxx,rgbx_channel);
-    split(gyy,rgby_channel);
-
-    split(I,output);
-
-    poisson_solver(rgb_channel[2],rgbx_channel[2], rgby_channel[2],output[2]);
-    poisson_solver(rgb_channel[1],rgbx_channel[1], rgby_channel[1],output[1]);
-    poisson_solver(rgb_channel[0],rgbx_channel[0], rgby_channel[0],output[0]);
-}
-
-void Cloning::evaluate(Mat &I, Mat &wmask, Mat &cloned)
-{
-    bitwise_not(wmask,wmask);
-
-    wmask.convertTo(smask1,CV_32FC1,1.0/255.0);
-    I.convertTo(grx32,CV_32FC3,1.0/255.0);
-    I.convertTo(gry32,CV_32FC3,1.0/255.0);
-
-    array_product(grx32,grx,smask1);
-    array_product(gry32,gry,smask1);
-
-    poisson(I,grx32,gry32,srx32,sry32);
-
-    merge(output,cloned);
-}
-
-void Cloning::normal_clone(Mat &I, Mat &mask, Mat &wmask, Mat &cloned, int num)
-{
-    int w = I.size().width;
-    int h = I.size().height;
-    int channel = I.channels();
-
-
-    initialization(I,mask,wmask);
-
-    if(num == 1)
-    {
-        array_product(srx32,sgx,smask);
-        array_product(sry32,sgy,smask);
-
-    }
-    else if(num == 2)
-    {
-
-        for(int i=0;i < h; i++)
-        {
-           for(int j=0; j < w; j++)
-            {
-                for(int c=0;c<channel;++c)
-                {
-                    if(abs(sgx.at<float>(i,j*channel+c) - sgy.at<float>(i,j*channel+c)) >
-                            abs(grx.at<float>(i,j*channel+c) - gry.at<float>(i,j*channel+c)))
-                    {
-
-                        srx32.at<float>(i,j*channel+c) = sgx.at<float>(i,j*channel+c)
-                            * smask.at<float>(i,j);
-                        sry32.at<float>(i,j*channel+c) = sgy.at<float>(i,j*channel+c)
-                            * smask.at<float>(i,j);
-                    }
-                    else
-                    {
-                        srx32.at<float>(i,j*channel+c) = grx.at<float>(i,j*channel+c)
-                            * smask.at<float>(i,j);
-                        sry32.at<float>(i,j*channel+c) = gry.at<float>(i,j*channel+c)
-                            * smask.at<float>(i,j);
-                    }
-                }
-            }
-        }
-
-    }
-    else if(num == 3)
-    {
-        Mat gray = Mat(mask.size(),CV_8UC1);
-        Mat gray8 = Mat(mask.size(),CV_8UC3);
-        cvtColor(mask, gray, COLOR_BGR2GRAY );
-        vector <Mat> temp;
-        split(I,temp);
-        gray.copyTo(temp[2]);
-        gray.copyTo(temp[1]);
-        gray.copyTo(temp[0]);
-
-        merge(temp,gray8);
-
-        getGradientx(gray8,sgx);
-        getGradienty(gray8,sgy);
-
-        array_product(srx32,sgx,smask);
-        array_product(sry32,sgy,smask);
-
-    }
-
-    evaluate(I,wmask,cloned);
-}
-
-void Cloning::local_color_change(Mat &I, Mat &mask, Mat &wmask, Mat &cloned, float red_mul=1.0,
-                                 float green_mul=1.0, float blue_mul=1.0)
-{
-    initialization(I,mask,wmask);
-
-    array_product(srx32,sgx,smask);
-    array_product(sry32,sgy,smask);
-    scalar_product(srx32,red_mul,green_mul,blue_mul);
-    scalar_product(sry32,red_mul,green_mul,blue_mul);
-
-    evaluate(I,wmask,cloned);
-}
-
-void Cloning::illum_change(Mat &I, Mat &mask, Mat &wmask, Mat &cloned, float alpha, float beta)
-{
-    initialization(I,mask,wmask);
-
-    array_product(srx32,sgx,smask);
-    array_product(sry32,sgy,smask);
-
-    Mat mag = Mat(I.size(),CV_32FC3);
-    magnitude(srx32,sry32,mag);
-
-    Mat multX, multY, multx_temp, multy_temp;
-
-    multiply(srx32,pow(alpha,beta),multX);
-    pow(mag,-1*beta, multx_temp);
-    multiply(multX,multx_temp,srx32);
-    patchNaNs(srx32);
-
-    multiply(sry32,pow(alpha,beta),multY);
-    pow(mag,-1*beta, multy_temp);
-    multiply(multY,multy_temp,sry32);
-    patchNaNs(sry32);
-
-    Mat zeroMask = (srx32 != 0);
-
-    srx32.copyTo(srx32, zeroMask);
-    sry32.copyTo(sry32, zeroMask);
-
-    evaluate(I,wmask,cloned);
-}
-
-void Cloning::texture_flatten(Mat &I, Mat &mask, Mat &wmask, double low_threshold,
-        double high_threshold, int kernel_size, Mat &cloned)
-{
-    initialization(I,mask,wmask);
-
-    Mat out = Mat(mask.size(),CV_8UC1);
-    Canny(mask,out,low_threshold,high_threshold,kernel_size);
-
-    Mat zeros(sgx.size(), CV_32FC3);
-    zeros.setTo(0);
-    Mat zerosMask = (out != 255);
-    zeros.copyTo(sgx, zerosMask);
-    zeros.copyTo(sgy, zerosMask);
-
-    array_product(srx32,sgx,smask);
-    array_product(sry32,sgy,smask);
-
-    evaluate(I,wmask,cloned);
+            std::vector<float> filter_X, filter_Y;
+    };
 }
+#endif
\ No newline at end of file
diff --git a/modules/photo/src/seamless_cloning_impl.cpp b/modules/photo/src/seamless_cloning_impl.cpp
new file mode 100644
index 0000000000..fe2751d143
--- /dev/null
+++ b/modules/photo/src/seamless_cloning_impl.cpp
@@ -0,0 +1,469 @@
+/*M///////////////////////////////////////////////////////////////////////////////////////
+//
+//  IMPORTANT: READ BEFORE DOWNLOADING, COPYING, INSTALLING OR USING.
+//
+//  By downloading, copying, installing or using the software you agree to this license.
+//  If you do not agree to this license, do not download, install,
+//  copy or use the software.
+//
+//
+//                           License Agreement
+//                For Open Source Computer Vision Library
+//
+// Copyright (C) 2013, OpenCV Foundation, all rights reserved.
+// Third party copyrights are property of their respective owners.
+//
+// Redistribution and use in source and binary forms, with or without modification,
+// are permitted provided that the following conditions are met:
+//
+//   * Redistribution's of source code must retain the above copyright notice,
+//     this list of conditions and the following disclaimer.
+//
+//   * Redistribution's in binary form must reproduce the above copyright notice,
+//     this list of conditions and the following disclaimer in the documentation
+//     and/or other materials provided with the distribution.
+//
+//   * The name of the copyright holders may not be used to endorse or promote products
+//     derived from this software without specific prior written permission.
+//
+// This software is provided by the copyright holders and contributors "as is" and
+// any express or implied warranties, including, but not limited to, the implied
+// warranties of merchantability and fitness for a particular purpose are disclaimed.
+// In no event shall the Intel Corporation or contributors be liable for any direct,
+// indirect, incidental, special, exemplary, or consequential damages
+// (including, but not limited to, procurement of substitute goods or services;
+// loss of use, data, or profits; or business interruption) however caused
+// and on any theory of liability, whether in contract, strict liability,
+// or tort (including negligence or otherwise) arising in any way out of
+// the use of this software, even if advised of the possibility of such damage.
+//
+//M*/
+
+#include "seamless_cloning.hpp"
+
+using namespace cv;
+using namespace std;
+
+
+void Cloning::computeGradientX( const Mat &img, Mat &gx)
+{
+    Mat kernel = Mat::zeros(1, 3, CV_8S);
+    kernel.at<char>(0,2) = 1;
+    kernel.at<char>(0,1) = -1;
+
+    if(img.channels() == 3)
+    {
+        filter2D(img, gx, CV_32F, kernel);
+    }
+    else if (img.channels() == 1)
+    {
+        Mat tmp[3];
+        for(int chan = 0 ; chan < 3 ; ++chan)
+        {
+            filter2D(img, tmp[chan], CV_32F, kernel);
+        }
+        merge(tmp, 3, gx);
+    }
+}
+
+void Cloning::computeGradientY( const Mat &img, Mat &gy)
+{
+    Mat kernel = Mat::zeros(3, 1, CV_8S);
+    kernel.at<char>(2,0) = 1;
+    kernel.at<char>(1,0) = -1;
+
+    if(img.channels() == 3)
+    {
+        filter2D(img, gy, CV_32F, kernel);
+    }
+    else if (img.channels() == 1)
+    {
+        Mat tmp[3];
+        for(int chan = 0 ; chan < 3 ; ++chan)
+        {
+            filter2D(img, tmp[chan], CV_32F, kernel);
+        }
+        merge(tmp, 3, gy);
+    }
+}
+
+void Cloning::computeLaplacianX( const Mat &img, Mat &laplacianX)
+{
+    Mat kernel = Mat::zeros(1, 3, CV_8S);
+    kernel.at<char>(0,0) = -1;
+    kernel.at<char>(0,1) = 1;
+    filter2D(img, laplacianX, CV_32F, kernel);
+}
+
+void Cloning::computeLaplacianY( const Mat &img, Mat &laplacianY)
+{
+    Mat kernel = Mat::zeros(3, 1, CV_8S);
+    kernel.at<char>(0,0) = -1;
+    kernel.at<char>(1,0) = 1;
+    filter2D(img, laplacianY, CV_32F, kernel);
+}
+
+void Cloning::dst(const Mat& src, Mat& dest, bool invert)
+{
+    Mat temp = Mat::zeros(src.rows, 2 * src.cols + 2, CV_32F);
+
+    int flag = invert ? DFT_ROWS + DFT_SCALE + DFT_INVERSE: DFT_ROWS;
+
+    src.copyTo(temp(Rect(1,0, src.cols, src.rows)));
+
+    for(int j = 0 ; j < src.rows ; ++j)
+    {
+        float * tempLinePtr = temp.ptr<float>(j);
+        const float * srcLinePtr = src.ptr<float>(j);
+        for(int i = 0 ; i < src.cols ; ++i)
+        {
+            tempLinePtr[src.cols + 2 + i] = - srcLinePtr[src.cols - 1 - i];
+        }
+    }
+
+    Mat planes[] = {temp, Mat::zeros(temp.size(), CV_32F)};
+    Mat complex;
+
+    merge(planes, 2, complex);
+    dft(complex, complex, flag);
+    split(complex, planes);
+    temp = Mat::zeros(src.cols, 2 * src.rows + 2, CV_32F);
+
+    for(int j = 0 ; j < src.cols ; ++j)
+    {
+        float * tempLinePtr = temp.ptr<float>(j);
+        for(int i = 0 ; i < src.rows ; ++i)
+        {
+            float val = planes[1].ptr<float>(i)[j + 1];
+            tempLinePtr[i + 1] = val;
+            tempLinePtr[temp.cols - 1 - i] = - val;
+        }
+    }
+
+    Mat planes2[] = {temp, Mat::zeros(temp.size(), CV_32F)};
+
+    merge(planes2, 2, complex);
+    dft(complex, complex, flag);
+    split(complex, planes2);
+
+    temp = planes2[1].t();
+    dest = Mat::zeros(src.size(), CV_32F);
+    temp(Rect( 0, 1, src.cols, src.rows)).copyTo(dest);
+}
+
+void Cloning::idst(const Mat& src, Mat& dest)
+{
+    dst(src, dest, true);
+}
+
+void Cloning::solve(const Mat &img, Mat& mod_diff, Mat &result)
+{
+    const int w = img.cols;
+    const int h = img.rows;
+
+    Mat res;
+    dst(mod_diff, res);
+
+    for(int j = 0 ; j < h-2; j++)
+    {
+        float * resLinePtr = res.ptr<float>(j);
+        for(int i = 0 ; i < w-2; i++)
+        {
+            resLinePtr[i] /= (filter_X[i] + filter_Y[j] - 4);
+        }
+    }
+
+    idst(res, mod_diff);
+
+    unsigned char *  resLinePtr = result.ptr<unsigned char>(0);
+    const unsigned char * imgLinePtr = img.ptr<unsigned char>(0);
+    const float * interpLinePtr = NULL;
+
+     //first col
+    for(int i = 0 ; i < w ; ++i)
+        result.ptr<unsigned char>(0)[i] = img.ptr<unsigned char>(0)[i];
+
+    for(int j = 1 ; j < h-1 ; ++j)
+    {
+        resLinePtr = result.ptr<unsigned char>(j);
+        imgLinePtr  = img.ptr<unsigned char>(j);
+        interpLinePtr = mod_diff.ptr<float>(j-1);
+
+        //first row
+        resLinePtr[0] = imgLinePtr[0];
+
+        for(int i = 1 ; i < w-1 ; ++i)
+        {
+            //saturate cast is not used here, because it behaves differently from the previous implementation
+            //most notable, saturate_cast rounds before truncating, here it's the opposite.
+            float value = interpLinePtr[i-1];
+            if(value < 0.)
+                resLinePtr[i] = 0;
+            else if (value > 255.0)
+                resLinePtr[i] = 255;
+            else
+                resLinePtr[i] = static_cast<unsigned char>(value);
+        }
+
+        //last row
+        resLinePtr[w-1] = imgLinePtr[w-1];
+    }
+
+    //last col
+    resLinePtr = result.ptr<unsigned char>(h-1);
+    imgLinePtr = img.ptr<unsigned char>(h-1);
+    for(int i = 0 ; i < w ; ++i)
+        resLinePtr[i] = imgLinePtr[i];
+}
+
+void Cloning::poissonSolver(const Mat &img, Mat &laplacianX , Mat &laplacianY, Mat &result)
+{
+    const int w = img.cols;
+    const int h = img.rows;
+
+    Mat lap = Mat(img.size(),CV_32FC1);
+
+    lap = laplacianX + laplacianY;
+
+    Mat bound = img.clone();
+
+    rectangle(bound, Point(1, 1), Point(img.cols-2, img.rows-2), Scalar::all(0), -1);
+    Mat boundary_points;
+    Laplacian(bound, boundary_points, CV_32F);
+
+    boundary_points = lap - boundary_points;
+
+    Mat mod_diff = boundary_points(Rect(1, 1, w-2, h-2));
+
+    solve(img,mod_diff,result);
+}
+
+void Cloning::initVariables(const Mat &destination, const Mat &binaryMask)
+{
+    destinationGradientX = Mat(destination.size(),CV_32FC3);
+    destinationGradientY = Mat(destination.size(),CV_32FC3);
+    patchGradientX = Mat(destination.size(),CV_32FC3);
+    patchGradientY = Mat(destination.size(),CV_32FC3);
+
+    binaryMaskFloat = Mat(binaryMask.size(),CV_32FC1);
+    binaryMaskFloatInverted = Mat(binaryMask.size(),CV_32FC1);
+
+    //init of the filters used in the dst
+    const int w = destination.cols;
+    filter_X.resize(w - 2);
+    for(int i = 0 ; i < w-2 ; ++i)
+        filter_X[i] = 2.0f * std::cos(static_cast<float>(CV_PI) * (i + 1) / (w - 1));
+
+    const int h  = destination.rows;
+    filter_Y.resize(h - 2);
+    for(int j = 0 ; j < h - 2 ; ++j)
+        filter_Y[j] = 2.0f * std::cos(static_cast<float>(CV_PI) * (j + 1) / (h - 1));
+}
+
+void Cloning::computeDerivatives(const Mat& destination, const Mat &patch, const Mat &binaryMask)
+{
+    initVariables(destination,binaryMask);
+
+    computeGradientX(destination,destinationGradientX);
+    computeGradientY(destination,destinationGradientY);
+
+    computeGradientX(patch,patchGradientX);
+    computeGradientY(patch,patchGradientY);
+
+    Mat Kernel(Size(3, 3), CV_8UC1);
+    Kernel.setTo(Scalar(1));
+    erode(binaryMask, binaryMask, Kernel, Point(-1,-1), 3);
+
+    binaryMask.convertTo(binaryMaskFloat,CV_32FC1,1.0/255.0);
+}
+
+void Cloning::scalarProduct(Mat mat, float r, float g, float b)
+{
+    vector <Mat> channels;
+    split(mat,channels);
+    multiply(channels[2],r,channels[2]);
+    multiply(channels[1],g,channels[1]);
+    multiply(channels[0],b,channels[0]);
+    merge(channels,mat);
+}
+
+void Cloning::arrayProduct(const cv::Mat& lhs, const cv::Mat& rhs, cv::Mat& result) const
+{
+    vector <Mat> lhs_channels;
+    vector <Mat> result_channels;
+
+    split(lhs,lhs_channels);
+    split(result,result_channels);
+
+    for(int chan = 0 ; chan < 3 ; ++chan)
+        multiply(lhs_channels[chan],rhs,result_channels[chan]);
+
+    merge(result_channels,result);
+}
+
+void Cloning::poisson(const Mat &destination)
+{
+    Mat laplacianX = Mat(destination.size(),CV_32FC3);
+    Mat laplacianY = Mat(destination.size(),CV_32FC3);
+
+    laplacianX = destinationGradientX + patchGradientX;
+    laplacianY = destinationGradientY + patchGradientY;
+
+    computeLaplacianX(laplacianX,laplacianX);
+    computeLaplacianY(laplacianY,laplacianY);
+
+    split(laplacianX,rgbx_channel);
+    split(laplacianY,rgby_channel);
+
+    split(destination,output);
+
+    for(int chan = 0 ; chan < 3 ; ++chan)
+    {
+        poissonSolver(output[chan], rgbx_channel[chan], rgby_channel[chan], output[chan]);
+    }
+}
+
+void Cloning::evaluate(const Mat &I, const Mat &wmask, const Mat &cloned)
+{
+    bitwise_not(wmask,wmask);
+
+    wmask.convertTo(binaryMaskFloatInverted,CV_32FC1,1.0/255.0);
+
+    arrayProduct(destinationGradientX,binaryMaskFloatInverted, destinationGradientX);
+    arrayProduct(destinationGradientY,binaryMaskFloatInverted, destinationGradientY);
+
+    poisson(I);
+
+    merge(output,cloned);
+}
+
+void Cloning::normalClone(const Mat &destination, const Mat &patch, const Mat &binaryMask, Mat &cloned, int flag)
+{
+    const int w = destination.cols;
+    const int h = destination.rows;
+    const int channel = destination.channels();
+    const int n_elem_in_line = w * channel;
+
+    computeDerivatives(destination,patch,binaryMask);
+
+    switch(flag)
+    {
+        case NORMAL_CLONE:
+            arrayProduct(patchGradientX,binaryMaskFloat, patchGradientX);
+            arrayProduct(patchGradientY,binaryMaskFloat, patchGradientY);
+            break;
+
+        case MIXED_CLONE:
+        {
+            AutoBuffer<int> maskIndices(n_elem_in_line);
+            for (int i = 0; i < n_elem_in_line; ++i)
+                maskIndices[i] = i / channel;
+
+            for(int i=0;i < h; i++)
+            {
+                float * patchXLinePtr = patchGradientX.ptr<float>(i);
+                float * patchYLinePtr = patchGradientY.ptr<float>(i);
+                const float * destinationXLinePtr = destinationGradientX.ptr<float>(i);
+                const float * destinationYLinePtr = destinationGradientY.ptr<float>(i);
+                const float * binaryMaskLinePtr = binaryMaskFloat.ptr<float>(i);
+
+                for(int j=0; j < n_elem_in_line; j++)
+                {
+                    int maskIndex = maskIndices[j];
+
+                    if(abs(patchXLinePtr[j] - patchYLinePtr[j]) >
+                       abs(destinationXLinePtr[j] - destinationYLinePtr[j]))
+                    {
+                        patchXLinePtr[j] *= binaryMaskLinePtr[maskIndex];
+                        patchYLinePtr[j] *= binaryMaskLinePtr[maskIndex];
+                    }
+                    else
+                    {
+                        patchXLinePtr[j] = destinationXLinePtr[j]
+                            * binaryMaskLinePtr[maskIndex];
+                        patchYLinePtr[j] = destinationYLinePtr[j]
+                            * binaryMaskLinePtr[maskIndex];
+                    }
+                }
+            }
+        }
+        break;
+
+        case MONOCHROME_TRANSFER:
+            Mat gray = Mat(patch.size(),CV_8UC1);
+            cvtColor(patch, gray, COLOR_BGR2GRAY );
+
+            computeGradientX(gray,patchGradientX);
+            computeGradientY(gray,patchGradientY);
+
+            arrayProduct(patchGradientX, binaryMaskFloat, patchGradientX);
+            arrayProduct(patchGradientY, binaryMaskFloat, patchGradientY);
+        break;
+
+    }
+
+    evaluate(destination,binaryMask,cloned);
+}
+
+void Cloning::localColorChange(Mat &I, Mat &mask, Mat &wmask, Mat &cloned, float red_mul=1.0,
+                                 float green_mul=1.0, float blue_mul=1.0)
+{
+    computeDerivatives(I,mask,wmask);
+
+    arrayProduct(patchGradientX,binaryMaskFloat, patchGradientX);
+    arrayProduct(patchGradientY,binaryMaskFloat, patchGradientY);
+    scalarProduct(patchGradientX,red_mul,green_mul,blue_mul);
+    scalarProduct(patchGradientY,red_mul,green_mul,blue_mul);
+
+    evaluate(I,wmask,cloned);
+}
+
+void Cloning::illuminationChange(Mat &I, Mat &mask, Mat &wmask, Mat &cloned, float alpha, float beta)
+{
+    computeDerivatives(I,mask,wmask);
+
+    arrayProduct(patchGradientX,binaryMaskFloat, patchGradientX);
+    arrayProduct(patchGradientY,binaryMaskFloat, patchGradientY);
+
+    Mat mag = Mat(I.size(),CV_32FC3);
+    magnitude(patchGradientX,patchGradientY,mag);
+
+    Mat multX, multY, multx_temp, multy_temp;
+
+    multiply(patchGradientX,pow(alpha,beta),multX);
+    pow(mag,-1*beta, multx_temp);
+    multiply(multX,multx_temp, patchGradientX);
+    patchNaNs(patchGradientX);
+
+    multiply(patchGradientY,pow(alpha,beta),multY);
+    pow(mag,-1*beta, multy_temp);
+    multiply(multY,multy_temp,patchGradientY);
+    patchNaNs(patchGradientY);
+
+    Mat zeroMask = (patchGradientX != 0);
+
+    patchGradientX.copyTo(patchGradientX, zeroMask);
+    patchGradientY.copyTo(patchGradientY, zeroMask);
+
+    evaluate(I,wmask,cloned);
+}
+
+void Cloning::textureFlatten(Mat &I, Mat &mask, Mat &wmask, float low_threshold,
+        float high_threshold, int kernel_size, Mat &cloned)
+{
+    computeDerivatives(I,mask,wmask);
+
+    Mat out = Mat(mask.size(),CV_8UC1);
+    Canny(mask,out,low_threshold,high_threshold,kernel_size);
+
+    Mat zeros(patchGradientX.size(), CV_32FC3);
+    zeros.setTo(0);
+    Mat zerosMask = (out != 255);
+    zeros.copyTo(patchGradientX, zerosMask);
+    zeros.copyTo(patchGradientY, zerosMask);
+
+    arrayProduct(patchGradientX,binaryMaskFloat, patchGradientX);
+    arrayProduct(patchGradientY,binaryMaskFloat, patchGradientY);
+
+    evaluate(I,wmask,cloned);
+}
diff --git a/modules/photo/test/test_cloning.cpp b/modules/photo/test/test_cloning.cpp
index 62fe243954..56d166205c 100644
--- a/modules/photo/test/test_cloning.cpp
+++ b/modules/photo/test/test_cloning.cpp
@@ -39,6 +39,15 @@
 //
 //M*/
 
+#define OUTPUT_SAVING 0
+#if OUTPUT_SAVING
+#define SAVE(x) std::vector<int> params;\
+                params.push_back(16);\
+                params.push_back(0);\
+                imwrite(folder + "output.png", x ,params);
+#else
+#define SAVE(x)
+#endif
 
 #include "test_precomp.hpp"
 #include "opencv2/photo.hpp"
@@ -47,6 +56,7 @@
 using namespace cv;
 using namespace std;
 
+static const double numerical_precision = 1000.;
 
 TEST(Photo_SeamlessClone_normal, regression)
 {
@@ -69,8 +79,13 @@ TEST(Photo_SeamlessClone_normal, regression)
     p.y = destination.size().height/2;
     seamlessClone(source, destination, mask, p, result, 1);
 
-    imwrite(folder + "cloned.png", result);
 
+    Mat reference = imread(folder + "reference.png");
+
+    SAVE(result);
+
+    double error = cvtest::norm(reference, result, NORM_L1);
+    EXPECT_LE(error, numerical_precision);
 }
 
 TEST(Photo_SeamlessClone_mixed, regression)
@@ -94,7 +109,11 @@ TEST(Photo_SeamlessClone_mixed, regression)
     p.y = destination.size().height/2;
     seamlessClone(source, destination, mask, p, result, 2);
 
-    imwrite(folder + "cloned.png", result);
+    SAVE(result);
+
+    Mat reference = imread(folder + "reference.png");
+    double error = cvtest::norm(reference, result, NORM_L1);
+    EXPECT_LE(error, numerical_precision);
 
 }
 
@@ -119,7 +138,11 @@ TEST(Photo_SeamlessClone_featureExchange, regression)
     p.y = destination.size().height/2;
     seamlessClone(source, destination, mask, p, result, 3);
 
-    imwrite(folder + "cloned.png", result);
+    SAVE(result);
+
+    Mat reference = imread(folder + "reference.png");
+    double error = cvtest::norm(reference, result, NORM_L1);
+    EXPECT_LE(error, numerical_precision);
 
 }
 
@@ -138,7 +161,11 @@ TEST(Photo_SeamlessClone_colorChange, regression)
     Mat result;
     colorChange(source, mask, result, 1.5, .5, .5);
 
-    imwrite(folder + "cloned.png", result);
+    SAVE(result);
+
+    Mat reference = imread(folder + "reference.png");
+    double error = cvtest::norm(reference, result, NORM_L1);
+    EXPECT_LE(error, numerical_precision);
 
 }
 
@@ -157,7 +184,11 @@ TEST(Photo_SeamlessClone_illuminationChange, regression)
     Mat result;
     illuminationChange(source, mask, result, 0.2f, 0.4f);
 
-    imwrite(folder + "cloned.png", result);
+    SAVE(result);
+
+    Mat reference = imread(folder + "reference.png");
+    double error = cvtest::norm(reference, result, NORM_L1);
+    EXPECT_LE(error, numerical_precision);
 
 }
 
@@ -176,6 +207,10 @@ TEST(Photo_SeamlessClone_textureFlattening, regression)
     Mat result;
     textureFlattening(source, mask, result, 30, 45, 3);
 
-    imwrite(folder + "cloned.png", result);
+    SAVE(result);
+
+    Mat reference = imread(folder + "reference.png");
+    double error = cvtest::norm(reference, result, NORM_L1);
+    EXPECT_LE(error, numerical_precision);
 
 }
diff --git a/modules/photo/test/test_decolor.cpp b/modules/photo/test/test_decolor.cpp
index bf21f37384..259f7afd10 100644
--- a/modules/photo/test/test_decolor.cpp
+++ b/modules/photo/test/test_decolor.cpp
@@ -47,6 +47,7 @@
 using namespace cv;
 using namespace std;
 
+static const double numerical_precision = 10.;
 
 TEST(Photo_Decolor, regression)
 {
@@ -61,7 +62,11 @@ TEST(Photo_Decolor, regression)
         Mat grayscale, color_boost;
         decolor(original, grayscale, color_boost);
 
-        imwrite(folder + "grayscale.png",grayscale);
-        imwrite(folder + "color_boost.png",color_boost);
+        Mat reference_grayscale = imread(folder + "grayscale_reference.png", 0 /* == grayscale image*/);
+        double error_grayscale = cvtest::norm(reference_grayscale, grayscale, NORM_L1);
+        EXPECT_LE(error_grayscale, numerical_precision);
 
+        Mat reference_boost = imread(folder + "boost_reference.png");
+        double error_boost = cvtest::norm(reference_boost, color_boost, NORM_L1);
+        EXPECT_LE(error_boost, numerical_precision);
 }
diff --git a/modules/photo/test/test_npr.cpp b/modules/photo/test/test_npr.cpp
index 4d1c027ba1..24f6f886ec 100755
--- a/modules/photo/test/test_npr.cpp
+++ b/modules/photo/test/test_npr.cpp
@@ -47,6 +47,7 @@
 using namespace cv;
 using namespace std;
 
+static const double numerical_precision = 100.;
 
 TEST(Photo_NPR_EdgePreserveSmoothing_RecursiveFilter, regression)
 {
@@ -60,8 +61,9 @@ TEST(Photo_NPR_EdgePreserveSmoothing_RecursiveFilter, regression)
     Mat result;
     edgePreservingFilter(source,result,1);
 
-    imwrite(folder + "smoothened_RF.png", result);
-
+    Mat reference = imread(folder + "smoothened_RF_reference.png");
+    double error = cvtest::norm(reference, result, NORM_L1);
+    EXPECT_LE(error, numerical_precision);
 }
 
 TEST(Photo_NPR_EdgePreserveSmoothing_NormConvFilter, regression)
@@ -76,7 +78,9 @@ TEST(Photo_NPR_EdgePreserveSmoothing_NormConvFilter, regression)
     Mat result;
     edgePreservingFilter(source,result,2);
 
-    imwrite(folder + "smoothened_NCF.png", result);
+    Mat reference = imread(folder + "smoothened_NCF_reference.png");
+    double error = cvtest::norm(reference, result, NORM_L1);
+    EXPECT_LE(error, numerical_precision);
 
 }
 
@@ -92,8 +96,9 @@ TEST(Photo_NPR_DetailEnhance, regression)
     Mat result;
     detailEnhance(source,result);
 
-    imwrite(folder + "detail_enhanced.png", result);
-
+    Mat reference = imread(folder + "detail_enhanced_reference.png");
+    double error = cvtest::norm(reference, result, NORM_L1);
+    EXPECT_LE(error, numerical_precision);
 }
 
 TEST(Photo_NPR_PencilSketch, regression)
@@ -105,12 +110,16 @@ TEST(Photo_NPR_PencilSketch, regression)
 
     ASSERT_FALSE(source.empty()) << "Could not load input image " << original_path;
 
-    Mat result,result1;
-    pencilSketch(source,result,result1, 10, 0.1f, 0.03f);
+    Mat pencil_result, color_pencil_result;
+    pencilSketch(source,pencil_result, color_pencil_result, 10, 0.1f, 0.03f);
 
-    imwrite(folder + "pencil_sketch.png", result);
-    imwrite(folder + "color_pencil_sketch.png", result1);
+    Mat pencil_reference = imread(folder + "pencil_sketch_reference.png", 0 /* == grayscale*/);
+    double pencil_error = norm(pencil_reference, pencil_result, NORM_L1);
+    EXPECT_LE(pencil_error, numerical_precision);
 
+    Mat color_pencil_reference = imread(folder + "color_pencil_sketch_reference.png");
+    double color_pencil_error = cvtest::norm(color_pencil_reference, color_pencil_result, NORM_L1);
+    EXPECT_LE(color_pencil_error, numerical_precision);
 }
 
 TEST(Photo_NPR_Stylization, regression)
@@ -125,6 +134,8 @@ TEST(Photo_NPR_Stylization, regression)
     Mat result;
     stylization(source,result);
 
-    imwrite(folder + "stylized.png", result);
+    Mat stylized_reference = imread(folder + "stylized_reference.png");
+    double stylized_error = cvtest::norm(stylized_reference, result, NORM_L1);
+    EXPECT_LE(stylized_error, numerical_precision);
 
 }
diff --git a/samples/cpp/cloning_gui.cpp b/samples/cpp/cloning_gui.cpp
index 07671ca8a6..668dd2a959 100644
--- a/samples/cpp/cloning_gui.cpp
+++ b/samples/cpp/cloning_gui.cpp
@@ -65,7 +65,7 @@ float alpha,beta;
 
 float red, green, blue;
 
-double low_t, high_t;
+float low_t, high_t;
 
 void source(int, int, int, int, void*);
 void destination(int, int, int, int, void*);
diff --git a/samples/cpp/tutorial_code/photo/seamless_cloning/cloning_gui.cpp b/samples/cpp/tutorial_code/photo/seamless_cloning/cloning_gui.cpp
index 6edeb3eadb..6ca79a7a04 100644
--- a/samples/cpp/tutorial_code/photo/seamless_cloning/cloning_gui.cpp
+++ b/samples/cpp/tutorial_code/photo/seamless_cloning/cloning_gui.cpp
@@ -64,7 +64,7 @@ float alpha,beta;
 
 float red, green, blue;
 
-double low_t, high_t;
+float low_t, high_t;
 
 void source(int, int, int, int, void*);
 void destination(int, int, int, int, void*);