实例介绍
【实例简介】
【实例截图】
【核心代码】
// ;stitchimage.cpp : implementation file
//
#include "stdafx.h"
#include "imagestitch.h"
#include "stitchimage.h"
#ifdef _DEBUG
#define new DEBUG_NEW
#undef THIS_FILE
static char THIS_FILE[] = __FILE__;
#endif
/////////////////////////////////////////////////////////////////////////////
// stitchimage dialog
stitchimage::stitchimage(CWnd* pParent /*=NULL*/)
: CDialog(stitchimage::IDD, pParent)
{
//{{AFX_DATA_INIT(stitchimage)
// NOTE: the ClassWizard will add member initialization here
//}}AFX_DATA_INIT
}
void stitchimage::DoDataExchange(CDataExchange* pDX)
{
CDialog::DoDataExchange(pDX);
//{{AFX_DATA_MAP(stitchimage)
// NOTE: the ClassWizard will add DDX and DDV calls here
//}}AFX_DATA_MAP
}
BEGIN_MESSAGE_MAP(stitchimage, CDialog)
//{{AFX_MSG_MAP(stitchimage)
ON_BN_CLICKED(IDC_openone, Onopenone)
ON_BN_CLICKED(IDC_opentwo, Onopentwo)
ON_BN_CLICKED(IDC_palnetocy1, Onpalnetocy1)
ON_BN_CLICKED(IDC_palnetocy2, Onpalnetocy2)
ON_BN_CLICKED(IDC_step11, Onstep11)
ON_BN_CLICKED(IDC_step12, Onstep12)
ON_BN_CLICKED(IDC_step21, Onstep21)
ON_BN_CLICKED(IDC_step22, Onstep22)
ON_BN_CLICKED(IDC_step31, Onstep31)
ON_BN_CLICKED(IDC_step32, Onstep32)
ON_BN_CLICKED(IDC_step41, Onstep41)
ON_BN_CLICKED(IDC_step42, Onstep42)
ON_BN_CLICKED(IDC_step51, Onstep51)
ON_BN_CLICKED(IDC_step52, Onstep52)
ON_BN_CLICKED(IDC_siftstitch, Onsiftstitch)
//}}AFX_MSG_MAP
END_MESSAGE_MAP()
/////////////////////////////////////////////////////////////////////////////
// stitchimage message handlers
void stitchimage::Onopenone()
{
// TODO: Add your control notification handler code here
CString filename;
char szFilter[] = "Jpg Files (*.jpg)|*.jpg|All Files (*.*)|*.*||";
CFileDialog fdlg(TRUE, "JPG", NULL, OFN_HIDEREADONLY | OFN_OVERWRITEPROMPT,szFilter);
if(fdlg.DoModal()!=IDOK)//程序一直再进程中
return;
filename=fdlg.GetPathName();
std::string tempname = (LPCSTR)CString(filename);
const char* tmp=tempname.c_str();
imageone=NULL;
imageone=cvLoadImage(tmp,CV_LOAD_IMAGE_ANYCOLOR);
IplImage* dst=NULL;
dst=cvCreateImage(cvGetSize(imageone),imageone->depth,imageone->nChannels);
cvCopy(imageone,dst,NULL);
float wscale, hscale;
int tempw,temph;
tempw=imageone->width;
temph=imageone->height;
wscale=350.1/tempw;
hscale=240.0/temph;
CvSize dst_cvsize;
dst_cvsize.width = imageone->width * wscale; //目标图像的宽为源图象宽的scale倍
dst_cvsize.height = imageone->height * hscale; //目标图像的高为源图象高的scale倍
dst = cvCreateImage( dst_cvsize, imageone->depth, imageone->nChannels); //构造目标图象
cvResize(imageone, dst, CV_INTER_LINEAR); //缩放源图像到目标图像
cvNamedWindow( "图像1", CV_WINDOW_AUTOSIZE ); //创建用于显示目标图像的窗口
cvShowImage( "图像1", dst ); //显示目标图像
cvMoveWindow( "图像1",0,0);
cvWaitKey(0);
cvReleaseImage(&dst); //释放目标图像占用的内存
cvDestroyWindow( "图像1" ); //销毁窗口“dst”
//cvReleaseImage(&image1);
}
void stitchimage::Onopentwo()
{
// TODO: Add your control notification handler code here
CString filename;
char szFilter[] = "Jpg Files (*.jpg)|*.jpg|All Files (*.*)|*.*||";
CFileDialog fdlg(TRUE, "JPG", NULL, OFN_HIDEREADONLY | OFN_OVERWRITEPROMPT,szFilter);
if(fdlg.DoModal()!=IDOK)//程序一直再进程中
return;
filename=fdlg.GetPathName();
std::string tempname = (LPCSTR)CString(filename);
const char* tmp=tempname.c_str();
imagetwo=NULL;
imagetwo=cvLoadImage(tmp,CV_LOAD_IMAGE_ANYCOLOR);
IplImage* dst=NULL;
dst=cvCreateImage(cvGetSize(imagetwo),imagetwo->depth,imagetwo->nChannels);
cvCopy(imagetwo,dst,NULL);
float wscale, hscale;
int tempw,temph;
tempw=imagetwo->width;
temph=imagetwo->height;
wscale=350.1/tempw;
hscale=240.0/temph;
CvSize dst_cvsize;
dst_cvsize.width = imagetwo->width * wscale; //目标图像的宽为源图象宽的scale倍
dst_cvsize.height = imagetwo->height * hscale; //目标图像的高为源图象高的scale倍
dst = cvCreateImage( dst_cvsize, imagetwo->depth, imagetwo->nChannels); //构造目标图象
cvResize(imagetwo, dst, CV_INTER_LINEAR); //缩放源图像到目标图像
cvNamedWindow( "图像2", CV_WINDOW_AUTOSIZE ); //创建用于显示目标图像的窗口
cvShowImage( "图像2", dst ); //显示目标图像
cvMoveWindow( "图像2",360,0);
cvWaitKey(0);
cvReleaseImage(&dst); //释放目标图像占用的内存
cvDestroyWindow( "图像2" ); //销毁窗口“dst”
//cvReleaseImage(&image2);
}
void stitchimage::Onpalnetocy1()
{
// TODO: Add your control notification handler code here
imageptocy11=aa.palnetocylinder1(imageone);
imageptocy1=cutrightimage(imageptocy11);
// cvSaveImage("D:/imageptocy11.jpg",imageptocy11);
IplImage* dst3=NULL;
dst3=cvCreateImage(cvGetSize(imageptocy1),imageptocy1->depth,imageptocy1->nChannels);
cvCopy(imageptocy1,dst3,NULL);
int widthnew,heightnew;
widthnew=imageptocy1->width;
heightnew=imageptocy1->height;
float wscale ;
float hscale ;
wscale=350.0/widthnew;
hscale=240.0/heightnew;
CvSize dst3_cvsize;
dst3_cvsize.width = imageptocy1->width * wscale; //目标图像的宽为源图象宽的scale倍
dst3_cvsize.height =imageptocy1->height * hscale; //目标图像的高为源图象高的scale倍
dst3 = cvCreateImage( dst3_cvsize, imageptocy1->depth, imageptocy1->nChannels); //构造目标图象
cvResize(imageptocy1, dst3, CV_INTER_LINEAR); //缩放源图像到目标图像
cvNamedWindow( "柱面投影1", CV_WINDOW_AUTOSIZE ); //创建用于显示目标图像的窗口
cvShowImage( "柱面投影1", dst3 ); //显示目标图像
cvMoveWindow( "柱面投影1",0,250);
cvWaitKey(-1);
cvReleaseImage(&dst3); //释放目标图像占用的内存
cvDestroyWindow( "柱面投影1" ); //销毁窗口“dst”
}
void stitchimage::Onpalnetocy2()
{
// TODO: Add your control notification handler code here
imageptocy22=aa.palnetocylinder1(imagetwo);
imageptocy2=cutleftimage(imageptocy22);
// cvSaveImage("D:/imageptocy22.jpg",imageptocy22);
IplImage* dst3=NULL;
dst3=cvCreateImage(cvGetSize(imageptocy2),imageptocy2->depth,imageptocy2->nChannels);
cvCopy(imageptocy2,dst3,NULL);
int widthnew,heightnew;
widthnew=imageptocy2->width;
heightnew=imageptocy2->height;
float wscale ;
float hscale ;
wscale=350.0/widthnew;
hscale=240.0/heightnew;
CvSize dst3_cvsize;
dst3_cvsize.width = imageptocy2->width * wscale; //目标图像的宽为源图象宽的scale倍
dst3_cvsize.height =imageptocy2->height * hscale; //目标图像的高为源图象高的scale倍
dst3 = cvCreateImage( dst3_cvsize, imageptocy2->depth, imageptocy2->nChannels); //构造目标图象
cvResize(imageptocy2, dst3, CV_INTER_LINEAR); //缩放源图像到目标图像
cvNamedWindow( "柱面投影2", CV_WINDOW_AUTOSIZE ); //创建用于显示目标图像的窗口
cvShowImage( "柱面投影2", dst3 ); //显示目标图像
cvMoveWindow( "柱面投影2",360,250);
cvWaitKey(-1);
cvReleaseImage(&dst3); //释放目标图像占用的内存
cvDestroyWindow( "柱面投影2" ); //销毁窗口“dst”
}
void stitchimage::Onstep11()
{
// TODO: Add your control notification handler code here
image1=NULL;
IplImage *grey_im1=NULL;
DoubleSizeImage1=NULL;
CvMat *image1Mat=NULL;
tempMat1=NULL;
//为图像分配内存
image1=cvCreateImage(cvSize(imageptocy1->width,imageptocy1->height),imageptocy1->depth,imageptocy1->nChannels);
grey_im1=cvCreateImage(cvSize(imageptocy1->width,imageptocy1->height),IPL_DEPTH_8U,1);
DoubleSizeImage1=cvCreateImage(cvSize(2*(imageptocy1->width),2*(imageptocy1->height)),IPL_DEPTH_8U,3);
//为图像阵列分配内存，假设两幅图像的大小相同，tempMat跟随image1的大小
image1Mat=cvCreateMat(imageptocy1->height,imageptocy1->width,CV_32FC1);
//转化成单通道图像再处理
cvCvtColor(imageptocy1,grey_im1,CV_BGR2GRAY);
//转换到Mat数据结构，图像操作使用的是浮点型操作
cvConvert(grey_im1,image1Mat);
//double t=(double)cvGetTickCount();
//图像归一化
cvConvertScale(image1Mat,image1Mat,1.0/255,0);
int dim;
dim=min(image1Mat->rows,image1Mat->cols);
//金字塔阶数
aa.numoctaves=(int) (log((double)dim) / log(2.0))-2;
aa.numoctaves=min(aa.numoctaves,MAXOCTAVES);
//step1预滤波除噪声，建立金字塔底层，底层为原图像的两倍
tempMat1=aa.ScaleInitImage(image1Mat);
// cvReleaseImage(& image1);
cvReleaseImage(& grey_im1);
cvReleaseMat(& image1Mat);
//*******平滑滤波后的图像保存主要为了调试试验
/*IplImage *tempimage=NULL;
tempimage=cvCreateImage(cvSize(tempMat1->width,tempMat1->height),IPL_DEPTH_8U,1);
cvConvertScale(tempMat1,tempMat1,255.0,0);
cvConvertScaleAbs(tempMat1,tempimage,1,0);
cvSaveImage("D:/tempimage1.jpg",tempimage);*/
//cvReleaseImage(& tempimage);
//*******end
}
void stitchimage::Onstep12()
{
// TODO: Add your control notification handler code here
image2=NULL;
IplImage *grey_im1=NULL;
DoubleSizeImage2=NULL;
CvMat *image1Mat=NULL;
tempMat2=NULL;
//为图像分配内存
image2=cvCreateImage(cvSize(imageptocy2->width,imageptocy2->height),imageptocy2->depth,imageptocy2->nChannels);
grey_im1=cvCreateImage(cvSize(imageptocy2->width,imageptocy2->height),IPL_DEPTH_8U,1);
DoubleSizeImage2=cvCreateImage(cvSize(2*(imageptocy2->width),2*(imageptocy2->height)),IPL_DEPTH_8U,3);
//为图像阵列分配内存，假设两幅图像的大小相同，tempMat跟随image1的大小
image1Mat=cvCreateMat(imageptocy2->height,imageptocy2->width,CV_32FC1);
//转化成单通道图像再处理
cvCvtColor(imageptocy2,grey_im1,CV_BGR2GRAY);
//转换到Mat数据结构，图像操作使用的是浮点型操作
cvConvert(grey_im1,image1Mat);
//double t=(double)cvGetTickCount();
//图像归一化
cvConvertScale(image1Mat,image1Mat,1.0/255,0);
int dim;
dim=min(image1Mat->rows,image1Mat->cols);
//金字塔阶数
aa.numoctaves=(int) (log((double)dim) / log(2.0))-2;
aa.numoctaves=min(aa.numoctaves,MAXOCTAVES);
//step1预滤波除噪声，建立金字塔底层，底层为原图像的两倍
tempMat2=aa.ScaleInitImage(image1Mat);
//cvReleaseImage(& image2);
cvReleaseImage(& grey_im1);
cvReleaseMat(& image1Mat);
//*******平滑滤波后的图像保存主要为了调试试验
/*IplImage *tempimage=NULL;
tempimage=cvCreateImage(cvSize(tempMat2->width,tempMat2->height),IPL_DEPTH_8U,1);
cvConvertScale(tempMat2,tempMat2,255.0,0);
cvConvertScaleAbs(tempMat2,tempimage,1,0);
cvSaveImage("D:/tempimage2.jpg",tempimage);*/
//cvReleaseImage(& tempimage);
//*******end
}
IplImage* stitchimage::cutrightimage(IplImage *image)
{
int wd,hd;
wd=image->width;
hd=image->height;
if (wd >= 1600 && hd >= 1200)
{
uchar * datain;
uchar * tempdata;
IplImage *temp=NULL;
CvSize temp_cvsize;
temp_cvsize.width=1550;
temp_cvsize.height=1200;
temp=cvCreateImage(temp_cvsize,image->depth,image->nChannels);
tempdata=(uchar *)temp->imageData;
datain=(uchar *)image->imageData;
int wd,hg;
wd=image->width;
hg=image->height;
for (int i=0;i<1200;i )
{
for (int j=0;j<1550;j )
{
for (int n=0;n<3;n )
{
tempdata[i*temp->widthStep j*temp->nChannels n]=datain[(hg/2-600 i)*image->widthStep (wd-1550 j)*image->nChannels n];
}
}
}
// cvSaveImage("D:\imageptocy1.jpg",temp);
return temp;
}
else
{
return image;
}
}
/*void stitchimage::cutrightimage(IplImage * image)
{
uchar * datain;
uchar * tempdata;
IplImage *temp=NULL;
CvSize temp_cvsize;
temp_cvsize.width=1550;
temp_cvsize.height=1200;
temp=cvCreateImage(temp_cvsize,image->depth,image->nChannels);
tempdata=(uchar *)temp->imageData;
datain=(uchar *)image->imageData;
int wd,hg;
wd=image->width;
hg=image->height;
for (int i=0;i<1200;i )
{
for (int j=0;j<1550;j )
{
for (int n=0;n<3;n )
{
tempdata[i*temp->widthStep j*temp->nChannels n]=datain[(hg/2-600 i)*image->widthStep (wd-1550 j)*image->nChannels n];
}
}
}
imageptocy1=cvCreateImage(cvGetSize(temp),temp->depth,temp->nChannels);
cvCopy(temp,imageptocy1,NULL);
cvReleaseImage(&temp);
}*/
IplImage* stitchimage::cutleftimage(IplImage *image)
{
int wd,hd;
wd=image->width;
hd=image->height;
if (wd >= 1600 && hd >= 1200)
{
uchar * datain;
uchar * tempdata;
IplImage *temp=NULL;
CvSize temp_cvsize;
temp_cvsize.width=1550;
temp_cvsize.height=1200;
temp=cvCreateImage(temp_cvsize,image->depth,image->nChannels);
tempdata=(uchar *)temp->imageData;
datain=(uchar *)image->imageData;
int wd,hg;
wd=image->width;
hg=image->height;
for (int i=0;i<1200;i )
{
for (int j=0;j<1550;j )
{
for (int n=0;n<3;n )
{
tempdata[i*temp->widthStep j*temp->nChannels n]=datain[(hg/2-600 i)*image->widthStep j*image->nChannels n];
}
}
}
// cvSaveImage("D:\imageptocy1.jpg",temp);
return temp;
}
else
{
return image;
}
}
/*void stitchimage::cutleftimage(IplImage * image)
{
uchar * datain;
uchar * tempdata;
IplImage *temp=NULL;
CvSize temp_cvsize;
temp_cvsize.width=1550;
temp_cvsize.height=1200;
temp=cvCreateImage(temp_cvsize,image->depth,image->nChannels);
tempdata=(uchar *)temp->imageData;
datain=(uchar *)image->imageData;
int wd,hg;
wd=image->width;
hg=image->height;
for (int i=0;i<1200;i )
{
for (int j=0;j<1550;j )
{
for (int n=0;n<3;n )
{
tempdata[i*temp->widthStep j*temp->nChannels n]=datain[(hg/2-600 i)*image->widthStep j*image->nChannels n];
}
}
}
imageptocy2=cvCreateImage(cvGetSize(temp),temp->depth,temp->nChannels);
cvCopy(temp,imageptocy2,NULL);
cvReleaseImage(&temp);
//cvSaveImage("D:\cutimage.jpg",imageptocy2);
}*/
ImageOctaves * stitchimage::BuildGaussianOctaves1(CvMat * image)
{
ImageOctaves *octaves;
CvMat *tempMat11;
CvMat *dst;
CvMat *temp1;
int i,j;
double k=pow(2,1.0/((float)SCALESPEROCTAVE));//SCALESPEROCTAVE=2
float preblur_sigma,initial_sigma,sigma1,sigma2,sigma,absolute_sigma,sigma_f;
//计算金字塔的阶梯数目
int dim=min(image->rows,image->cols);
//金字塔阶数
int numoctaves=(int)(log((double) dim) / log(2.0))-2;
//限定金字塔的阶梯数
numoctaves=min(numoctaves,MAXOCTAVES);//MAXOCTAVES=4
//为高斯金字塔和DOG金字塔分配内存
octaves=(ImageOctaves*) malloc( numoctaves * sizeof(ImageOctaves));
DOGoctaves1=(ImageOctaves*) malloc( numoctaves * sizeof(ImageOctaves));
//初始化原图像
tempMat11=cvCloneMat(image);
//preblur_sigma = 1.0;//sqrt(2 - 4*INITSIGMA*INITSIGMA),INITSIGMA=0.5;
initial_sigma = sqrt(2);//sqrt( (4*INITSIGMA*INITSIGMA) preblur_sigma * preblur_sigma );
//initial_sigma = sqrt(SIGMA * SIGMA - INITSIGMA * INITSIGMA * 4);
//在每一阶金字塔图像中建立不同的尺度图像
for (i=0;i<numoctaves;i )
{
//首先建立金字塔每一阶梯的最底层，其中0阶梯的最底层已经建立好
//为各个阶梯分配内存
octaves[i].Octave=(ImageLevels*) malloc((SCALESPEROCTAVE 3) * sizeof(ImageLevels));
DOGoctaves1[i].Octave=(ImageLevels*) malloc((SCALESPEROCTAVE 2) * sizeof(ImageLevels));
//存储各个阶梯的最底层
(octaves[i].Octave)[0].Level=tempMat11;
octaves[i].col=tempMat11->cols;
octaves[i].row=tempMat11->rows;
DOGoctaves1[i].col=tempMat11->cols;
DOGoctaves1[i].row=tempMat11->rows;
if (DOUBLE_BASE_IMAGE_SIZE)//DOUBLE_BASE_IMAGE_SIZE=1
{
octaves[i].subsample=pow(2,i);
}
else
{
octaves[i].subsample=pow(2,i);
}
if (i==0)
{
(octaves[0].Octave)[0].levelsigma=initial_sigma;
(octaves[0].Octave)[0].absolute_sigma = initial_sigma;
}
else
{
(octaves[i].Octave)[0].levelsigma=(octaves[i-1].Octave)[SCALESPEROCTAVE].levelsigma;
(octaves[i].Octave)[0].absolute_sigma=(octaves[i-1].Octave)[SCALESPEROCTAVE].absolute_sigma;
}
sigma=initial_sigma;//sqrt(2)
//建立本阶梯其他层图像
for (j=1;j<SCALESPEROCTAVE 3;j )
{
dst=cvCreateMat(tempMat11->rows,tempMat11->cols,CV_32FC1);//用于存储高斯层
temp1=cvCreateMat(tempMat11->rows,tempMat11->cols,CV_32FC1);//用于存储DOG层
// if(i!=0)
// {
// sigma1 = pow(k, j - 1) * ((octaves[i-1].Octave)[j-1].levelsigma);
// sigma2 = pow(k, j) * ((octaves[i].Octave)[j-1].levelsigma);
// sigma = sqrt(sigma2*sigma2 - sigma1*sigma1);
sigma_f= sqrt(k*k-1)*sigma;//k=sqrt(2),sigma=sqrt(2),sigma_f=sqrt(2)
// }
// else
// {
// sigma = sqrt(SIGMA * SIGMA - INITSIGMA * INITSIGMA * 4)*pow(k,j);
// }
sigma = k*sigma;//sigma=2
absolute_sigma=sigma*(octaves[i].subsample);//octaves[0].subsample=0.5
(octaves[i].Octave)[j].levelsigma=sigma;
(octaves[i].Octave)[j].absolute_sigma=absolute_sigma;
//产生高斯层
int length=aa.BlurImage((octaves[i].Octave)[j-1].Level,dst,sigma_f);//相应尺度
(octaves[i].Octave)[j].levelsigmalength=length;//返回尺度大小
(octaves[i].Octave)[j].Level=dst;//
//产生DOG层
cvSub(((octaves[i].Octave)[j]).Level,((octaves[i].Octave)[j-1]).Level,temp1,0);
cvAbsDiff(((octaves[i].Octave)[j]).Level,((octaves[i].Octave)[j-1]).Level,temp1);
((DOGoctaves1[i].Octave)[j-1]).Level=temp1;
}
tempMat11=aa.halfSizeImage(( (octaves[i].Octave)[SCALESPEROCTAVE].Level ));
/* IplImage *mosaic11=NULL;
mosaic11=cvCreateImage(cvSize(dst->width,dst->height),IPL_DEPTH_8U,1);
cvConvertScale(dst,dst,255.0,0);
cvConvertScaleAbs(dst,mosaic11,1,0);
cvSaveImage("D:\GaussianPyramid of me.jpg",mosaic11);*/
}
cvReleaseMat(&tempMat11);
cvReleaseMat(&dst);
// cvReleaseMat(&temp1);
return octaves;
}
ImageOctaves * stitchimage::BuildGaussianOctaves2(CvMat * image)
{
ImageOctaves *octaves;
CvMat *tempMat11;
CvMat *dst;
CvMat *temp1;
int i,j;
double k=pow(2,1.0/((float)SCALESPEROCTAVE));//SCALESPEROCTAVE=2
float preblur_sigma,initial_sigma,sigma1,sigma2,sigma,absolute_sigma,sigma_f;
//计算金字塔的阶梯数目
int dim=min(image->rows,image->cols);
//金字塔阶数
int numoctaves=(int)(log((double) dim) / log(2.0))-2;
//限定金字塔的阶梯数
numoctaves=min(numoctaves,MAXOCTAVES);//MAXOCTAVES=4
//为高斯金字塔和DOG金字塔分配内存
octaves=(ImageOctaves*) malloc( numoctaves * sizeof(ImageOctaves));
DOGoctaves2=(ImageOctaves*) malloc( numoctaves * sizeof(ImageOctaves));
//初始化原图像
tempMat11=cvCloneMat(image);
//preblur_sigma = 1.0;//sqrt(2 - 4*INITSIGMA*INITSIGMA),INITSIGMA=0.5;
initial_sigma = sqrt(2);//sqrt( (4*INITSIGMA*INITSIGMA) preblur_sigma * preblur_sigma );
//initial_sigma = sqrt(SIGMA * SIGMA - INITSIGMA * INITSIGMA * 4);
//在每一阶金字塔图像中建立不同的尺度图像
for (i=0;i<numoctaves;i )
{
//首先建立金字塔每一阶梯的最底层，其中0阶梯的最底层已经建立好
//为各个阶梯分配内存
octaves[i].Octave=(ImageLevels*) malloc((SCALESPEROCTAVE 3) * sizeof(ImageLevels));
DOGoctaves2[i].Octave=(ImageLevels*) malloc((SCALESPEROCTAVE 2) * sizeof(ImageLevels));
//存储各个阶梯的最底层
(octaves[i].Octave)[0].Level=tempMat11;
octaves[i].col=tempMat11->cols;
octaves[i].row=tempMat11->rows;
DOGoctaves2[i].col=tempMat11->cols;
DOGoctaves2[i].row=tempMat11->rows;
if (DOUBLE_BASE_IMAGE_SIZE)//DOUBLE_BASE_IMAGE_SIZE=1
{
octaves[i].subsample=pow(2,i);
}
else
{
octaves[i].subsample=pow(2,i);
}
if (i==0)
{
(octaves[0].Octave)[0].levelsigma=initial_sigma;
(octaves[0].Octave)[0].absolute_sigma = initial_sigma;
}
else
{
(octaves[i].Octave)[0].levelsigma=(octaves[i-1].Octave)[SCALESPEROCTAVE].levelsigma;
(octaves[i].Octave)[0].absolute_sigma=(octaves[i-1].Octave)[SCALESPEROCTAVE].absolute_sigma;
}
sigma=initial_sigma;//sqrt(2)
//建立本阶梯其他层图像
for (j=1;j<SCALESPEROCTAVE 3;j )
{
dst=cvCreateMat(tempMat11->rows,tempMat11->cols,CV_32FC1);//用于存储高斯层
temp1=cvCreateMat(tempMat11->rows,tempMat11->cols,CV_32FC1);//用于存储DOG层
// if(i!=0)
// {
// sigma1 = pow(k, j - 1) * ((octaves[i-1].Octave)[j-1].levelsigma);
// sigma2 = pow(k, j) * ((octaves[i].Octave)[j-1].levelsigma);
// sigma = sqrt(sigma2*sigma2 - sigma1*sigma1);
sigma_f= sqrt(k*k-1)*sigma;//k=sqrt(2),sigma=sqrt(2),sigma_f=sqrt(2)
// }
// else
// {
// sigma = sqrt(SIGMA * SIGMA - INITSIGMA * INITSIGMA * 4)*pow(k,j);
// }
sigma = k*sigma;//sigma=2
absolute_sigma=sigma*(octaves[i].subsample);//octaves[0].subsample=0.5
(octaves[i].Octave)[j].levelsigma=sigma;
(octaves[i].Octave)[j].absolute_sigma=absolute_sigma;
//产生高斯层
int length=aa.BlurImage((octaves[i].Octave)[j-1].Level,dst,sigma_f);//相应尺度
(octaves[i].Octave)[j].levelsigmalength=length;//返回尺度大小
(octaves[i].Octave)[j].Level=dst;//
//产生DOG层
cvSub(((octaves[i].Octave)[j]).Level,((octaves[i].Octave)[j-1]).Level,temp1,0);
cvAbsDiff(((octaves[i].Octave)[j]).Level,((octaves[i].Octave)[j-1]).Level,temp1);
((DOGoctaves2[i].Octave)[j-1]).Level=temp1;
}
tempMat11=aa.halfSizeImage(( (octaves[i].Octave)[SCALESPEROCTAVE].Level ));
/* IplImage *mosaic11=NULL;
mosaic11=cvCreateImage(cvSize(dst->width,dst->height),IPL_DEPTH_8U,1);
cvConvertScale(dst,dst,255.0,0);
cvConvertScaleAbs(dst,mosaic11,1,0);
cvSaveImage("D:\GaussianPyramid of me.jpg",mosaic11);*/
}
cvReleaseMat(&tempMat11);
cvReleaseMat(&dst);
// cvReleaseMat(&temp1);//释放就出错
return octaves;
}
void stitchimage::Onstep21()
{
// TODO: Add your control notification handler code here
//step2建立Guassian金字塔和DOG金字塔
Gaussianpyr1=BuildGaussianOctaves1(tempMat1);
//显示高斯图像
/* CvMat *mosaicHorizen1=NULL;
CvMat *mosaicHorizen2=NULL;
CvMat *mosaicVertical1=NULL;
#define ImLevels(OCTAVE,LEVEL,ROW,COL) ((float *)(Gaussianpyr1[(OCTAVE)].Octave[(LEVEL)].Level->data.fl Gaussianpyr1[(OCTAVE)].Octave[(LEVEL)].Level->step/sizeof(float) *(ROW)))[(COL)]
for (int i=0;i<aa.numoctaves;i )
{
if (i==0)
{
mosaicHorizen1=aa.MosaicHorizen((Gaussianpyr1[0].Octave)[0].Level,(Gaussianpyr1[0].Octave)[1].Level);
for(int j=2;j<SCALESPEROCTAVE 3;j )
mosaicHorizen1=aa.MosaicHorizen( mosaicHorizen1, (Gaussianpyr1[0].Octave)[j].Level);
for(j=0;j<NUMSIZE;j )
mosaicHorizen1=aa.halfSizeImage(mosaicHorizen1);
}
else if (i==1)
{
mosaicHorizen2=aa.MosaicHorizen((Gaussianpyr1[1].Octave)[0].Level, (Gaussianpyr1[1].Octave)[1].Level);
for(int j=2;j<SCALESPEROCTAVE 3;j )
mosaicHorizen2=aa.MosaicHorizen(mosaicHorizen2, (Gaussianpyr1[1].Octave)[j].Level );
for(j=0;j<NUMSIZE;j )
mosaicHorizen2=aa.halfSizeImage(mosaicHorizen2);
mosaicVertical1=aa.MosaicVertical( mosaicHorizen1, mosaicHorizen2 );
}
else
{
mosaicHorizen1=aa.MosaicHorizen( (Gaussianpyr1[i].Octave)[0].Level, (Gaussianpyr1[i].Octave)[1].Level );
for (int j=2;j<SCALESPEROCTAVE 3;j )
mosaicHorizen1=aa.MosaicHorizen( mosaicHorizen1, (Gaussianpyr1[i].Octave)[j].Level );
for ( j=0;j<NUMSIZE;j )
mosaicHorizen1=aa.halfSizeImage(mosaicHorizen1);
mosaicVertical1=aa.MosaicVertical( mosaicVertical1, mosaicHorizen1 );
}
}
IplImage *mosaic1=NULL;
mosaic1=cvCreateImage(cvSize(mosaicVertical1->width,mosaicVertical1->height),IPL_DEPTH_8U,1);
cvConvertScale(mosaicVertical1,mosaicVertical1,255.0,0);
cvConvertScaleAbs(mosaicVertical1,mosaic1,1,0);
cvSaveImage("D:\GaussianPyramid of me.jpg",mosaic1);
cvNamedWindow("高斯金字塔",1);
cvShowImage("高斯金字塔", mosaic1);
cvWaitKey(0);
cvDestroyWindow("mosaic1");
cvReleaseImage(& mosaic1);
//显示DOG金字塔
for ( i=0; i<aa.numoctaves;i )
{
if (i==0)
{
mosaicHorizen1=aa.MosaicHorizen( (DOGoctaves1[0].Octave)[0].Level, (DOGoctaves1[0].Octave)[1].Level );
for (int j=2;j<SCALESPEROCTAVE 2;j )
mosaicHorizen1=aa.MosaicHorizen( mosaicHorizen1, (DOGoctaves1[0].Octave)[j].Level );
for ( j=0;j<NUMSIZE;j )
mosaicHorizen1=aa.halfSizeImage(mosaicHorizen1);
}
else if (i==1)
{
mosaicHorizen2=aa.MosaicHorizen( (DOGoctaves1[1].Octave)[0].Level, (DOGoctaves1[1].Octave)[1].Level );
for (int j=2;j<SCALESPEROCTAVE 2;j )
mosaicHorizen2=aa.MosaicHorizen( mosaicHorizen2, (DOGoctaves1[1].Octave)[j].Level );
for ( j=0;j<NUMSIZE;j )
mosaicHorizen2=aa.halfSizeImage(mosaicHorizen2);
mosaicVertical1=aa.MosaicVertical( mosaicHorizen1, mosaicHorizen2 );
}
else
{
mosaicHorizen1=aa.MosaicHorizen( (DOGoctaves1[i].Octave)[0].Level, (DOGoctaves1[i].Octave)[1].Level );
for (int j=2;j<SCALESPEROCTAVE 2;j )
mosaicHorizen1=aa.MosaicHorizen( mosaicHorizen1, (DOGoctaves1[i].Octave)[j].Level );
for ( j=0;j<NUMSIZE;j )
mosaicHorizen1=aa.halfSizeImage(mosaicHorizen1);
mosaicVertical1=aa.MosaicVertical( mosaicVertical1, mosaicHorizen1 );
}
}
//考虑到DOG金字塔各层图像都会有正负，所以，必须寻找最负的，以将所有图像抬高一个台阶去显示
double min_val=0;
double max_val=0;
cvMinMaxLoc(mosaicVertical1,&min_val,&max_val,NULL,NULL,NULL);
if (min_val<0.0)
{
cvAddS(mosaicVertical1,cvScalarAll((-1)*min_val),mosaicVertical1,NULL);
}
IplImage *mosaic2=NULL;
mosaic2 = cvCreateImage(cvSize(mosaicVertical1->width, mosaicVertical1->height), IPL_DEPTH_8U,1);
cvConvertScale( mosaicVertical1, mosaicVertical1, 255.0/(max_val-min_val), 0 );
cvConvertScaleAbs( mosaicVertical1, mosaic2, 1, 0 );
cvSaveImage("DOGPyramid of me.jpg",mosaic2);
cvNamedWindow("mosaic1",1);
cvShowImage("mosaic1", mosaic2);
cvWaitKey(0);*/
}
void stitchimage::Onstep22()
{
// TODO: Add your control notification handler code here
Gaussianpyr2=BuildGaussianOctaves2(tempMat2);
//显示高斯图像
/* CvMat *mosaicHorizen1=NULL;
CvMat *mosaicHorizen2=NULL;
CvMat *mosaicVertical1=NULL;
#define ImLevels(OCTAVE,LEVEL,ROW,COL) ((float *)(Gaussianpyr2[(OCTAVE)].Octave[(LEVEL)].Level->data.fl Gaussianpyr2[(OCTAVE)].Octave[(LEVEL)].Level->step/sizeof(float) *(ROW)))[(COL)]
for (int i=0;i<aa.numoctaves;i )
{
if (i==0)
{
mosaicHorizen1=aa.MosaicHorizen((Gaussianpyr2[0].Octave)[0].Level,(Gaussianpyr2[0].Octave)[1].Level);
for(int j=2;j<SCALESPEROCTAVE 3;j )
mosaicHorizen1=aa.MosaicHorizen( mosaicHorizen1, (Gaussianpyr2[0].Octave)[j].Level);
for(j=0;j<NUMSIZE;j )
mosaicHorizen1=aa.halfSizeImage(mosaicHorizen1);
}
else if (i==1)
{
mosaicHorizen2=aa.MosaicHorizen((Gaussianpyr2[1].Octave)[0].Level, (Gaussianpyr2[1].Octave)[1].Level);
for(int j=2;j<SCALESPEROCTAVE 3;j )
mosaicHorizen2=aa.MosaicHorizen(mosaicHorizen2, (Gaussianpyr2[1].Octave)[j].Level );
for(j=0;j<NUMSIZE;j )
mosaicHorizen2=aa.halfSizeImage(mosaicHorizen2);
mosaicVertical1=aa.MosaicVertical( mosaicHorizen1, mosaicHorizen2 );
}
else
{
mosaicHorizen1=aa.MosaicHorizen( (Gaussianpyr2[i].Octave)[0].Level, (Gaussianpyr2[i].Octave)[1].Level );
for (int j=2;j<SCALESPEROCTAVE 3;j )
mosaicHorizen1=aa.MosaicHorizen( mosaicHorizen1, (Gaussianpyr2[i].Octave)[j].Level );
for ( j=0;j<NUMSIZE;j )
mosaicHorizen1=aa.halfSizeImage(mosaicHorizen1);
mosaicVertical1=aa.MosaicVertical( mosaicVertical1, mosaicHorizen1 );
}
}
IplImage *mosaic1=NULL;
mosaic1=cvCreateImage(cvSize(mosaicVertical1->width,mosaicVertical1->height),IPL_DEPTH_8U,1);
cvConvertScale(mosaicVertical1,mosaicVertical1,255.0,0);
cvConvertScaleAbs(mosaicVertical1,mosaic1,1,0);
cvSaveImage("D:\GaussianPyramid of me.jpg",mosaic1);
cvNamedWindow("高斯金字塔",1);
cvShowImage("高斯金字塔", mosaic1);
cvWaitKey(0);
cvDestroyWindow("mosaic1");
cvReleaseImage(& mosaic1);
//显示DOG金字塔
for ( i=0; i<aa.numoctaves;i )
{
if (i==0)
{
mosaicHorizen1=aa.MosaicHorizen( (DOGoctaves2[0].Octave)[0].Level, (DOGoctaves2[0].Octave)[1].Level );
for (int j=2;j<SCALESPEROCTAVE 2;j )
mosaicHorizen1=aa.MosaicHorizen( mosaicHorizen1, (DOGoctaves2[0].Octave)[j].Level );
for ( j=0;j<NUMSIZE;j )
mosaicHorizen1=aa.halfSizeImage(mosaicHorizen1);
}
else if (i==1)
{
mosaicHorizen2=aa.MosaicHorizen( (DOGoctaves2[1].Octave)[0].Level, (DOGoctaves2[1].Octave)[1].Level );
for (int j=2;j<SCALESPEROCTAVE 2;j )
mosaicHorizen2=aa.MosaicHorizen( mosaicHorizen2, (DOGoctaves2[1].Octave)[j].Level );
for ( j=0;j<NUMSIZE;j )
mosaicHorizen2=aa.halfSizeImage(mosaicHorizen2);
mosaicVertical1=aa.MosaicVertical( mosaicHorizen1, mosaicHorizen2 );
}
else
{
mosaicHorizen1=aa.MosaicHorizen( (DOGoctaves2[i].Octave)[0].Level, (DOGoctaves2[i].Octave)[1].Level );
for (int j=2;j<SCALESPEROCTAVE 2;j )
mosaicHorizen1=aa.MosaicHorizen( mosaicHorizen1, (DOGoctaves2[i].Octave)[j].Level );
for ( j=0;j<NUMSIZE;j )
mosaicHorizen1=aa.halfSizeImage(mosaicHorizen1);
mosaicVertical1=aa.MosaicVertical( mosaicVertical1, mosaicHorizen1 );
}
}
//考虑到DOG金字塔各层图像都会有正负，所以，必须寻找最负的，以将所有图像抬高一个台阶去显示
double min_val=0;
double max_val=0;
cvMinMaxLoc(mosaicVertical1,&min_val,&max_val,NULL,NULL,NULL);
if (min_val<0.0)
{
cvAddS(mosaicVertical1,cvScalarAll((-1)*min_val),mosaicVertical1,NULL);
}
IplImage *mosaic2=NULL;
mosaic2 = cvCreateImage(cvSize(mosaicVertical1->width, mosaicVertical1->height), IPL_DEPTH_8U,1);
cvConvertScale( mosaicVertical1, mosaicVertical1, 255.0/(max_val-min_val), 0 );
cvConvertScaleAbs( mosaicVertical1, mosaic2, 1, 0 );
cvSaveImage("DOGPyramid of me.jpg",mosaic2);
cvNamedWindow("mosaic1",1);
cvShowImage("mosaic1", mosaic2);
cvWaitKey(0);*/
}
int stitchimage::DetectKeypoint1(int numoctaves,ImageOctaves *Gaussianpyr1)
{
aa.keypoints1=NULL;
//计算用于DOG极值点检测的主曲率比的阈值
double curvature_threshold=((CURVATURE_THRESHOLD 1)*(CURVATURE_THRESHOLD 1))/CURVATURE_THRESHOLD;
#define ImLevels(OCTAVE,LEVEL,ROW,COL) ((float *)(DOGoctaves1[(OCTAVE)].Octave[(LEVEL)].Level->data.fl DOGoctaves1[(OCTAVE)].Octave[(LEVEL)].Level->step/sizeof(float) *(ROW)))[(COL)]
int keypoint_count=0;
for (int i=0;i<numoctaves;i )
{
for (int j=1;j<SCALESPEROCTAVE 1;j )//取中间的scaleperoctave个层
{
//在图像的有效区域内寻找具有显著性特征的局部最大值
int dim=(int)(0.5 ((Gaussianpyr1[i].Octave)[j].levelsigmalength) 0.5);
for (int m=dim;m<((DOGoctaves1[i].row)-dim);m )
{
for (int n=dim;n<((DOGoctaves1[i].col)-dim);n )
{
if (ImLevels(i,j,m,n)!=0.0)//first 非零
{
float inf_val=ImLevels(i,j,m,n);
/* if(( (inf_val <= 50 ImLevels(i,j-1,m-1,n-1))&&
(inf_val <= 50 ImLevels(i,j-1,m ,n-1))&&
(inf_val <= 50 ImLevels(i,j-1,m 1,n-1))&&
(inf_val <= 50 ImLevels(i,j-1,m-1,n ))&&
(inf_val <= 50 ImLevels(i,j-1,m ,n ))&&
(inf_val <= 50 ImLevels(i,j-1,m 1,n ))&&
(inf_val <= 50 ImLevels(i,j-1,m-1,n 1))&&
(inf_val <= 50 ImLevels(i,j-1,m ,n 1))&&
(inf_val <= 50 ImLevels(i,j-1,m 1,n 1))&& //底层的小尺度9
(inf_val <= 50 ImLevels(i,j,m-1,n-1))&&
(inf_val <= 50 ImLevels(i,j,m ,n-1))&&
(inf_val <= 50 ImLevels(i,j,m 1,n-1))&&
(inf_val <= 50 ImLevels(i,j,m-1,n ))&&
(inf_val <= 50 ImLevels(i,j,m 1,n ))&&
(inf_val <= 50 ImLevels(i,j,m-1,n 1))&&
(inf_val <= 50 ImLevels(i,j,m ,n 1))&&
(inf_val <= 50 ImLevels(i,j,m 1,n 1))&& //当前层8
(inf_val <= 50 ImLevels(i,j 1,m-1,n-1))&&
(inf_val <= 50 ImLevels(i,j 1,m ,n-1))&&
(inf_val <= 50 ImLevels(i,j 1,m 1,n-1))&&
(inf_val <= 50 ImLevels(i,j 1,m-1,n ))&&
(inf_val <= 50 ImLevels(i,j 1,m ,n ))&&
(inf_val <= 50 ImLevels(i,j 1,m 1,n ))&&
(inf_val <= 50 ImLevels(i,j 1,m-1,n 1))&&
(inf_val <= 50 ImLevels(i,j 1,m ,n 1))&&
(inf_val <= 50 ImLevels(i,j 1,m 1,n 1)) //下一层大尺度9
) ||
( (inf_val >= ImLevels(i,j-1,m-1,n-1)-5)&&
(inf_val >= ImLevels(i,j-1,m ,n-1)-5)&&
(inf_val >= ImLevels(i,j-1,m 1,n-1)-5)&&
(inf_val >= ImLevels(i,j-1,m-1,n )-5)&&
(inf_val >= ImLevels(i,j-1,m ,n )-5)&&
(inf_val >= ImLevels(i,j-1,m 1,n )-5)&&
(inf_val >= ImLevels(i,j-1,m-1,n 1)-5)&&
(inf_val >= ImLevels(i,j-1,m ,n 1)-5)&&
(inf_val >= ImLevels(i,j-1,m 1,n 1)-5)&&
(inf_val >= ImLevels(i,j,m-1,n-1)-5)&&
(inf_val >= ImLevels(i,j,m ,n-1)-5)&&
(inf_val >= ImLevels(i,j,m 1,n-1)-5)&&
(inf_val >= ImLevels(i,j,m-1,n )-5)&&
(inf_val >= ImLevels(i,j,m 1,n )-5)&&
(inf_val >= ImLevels(i,j,m-1,n 1)-5)&&
(inf_val >= ImLevels(i,j,m ,n 1)-5)&&
(inf_val >= ImLevels(i,j,m 1,n 1)-5)&&
(inf_val >= ImLevels(i,j 1,m-1,n-1)-5)&&
(inf_val >= ImLevels(i,j 1,m ,n-1)-5)&&
(inf_val >= ImLevels(i,j 1,m 1,n-1)-5)&&
(inf_val >= ImLevels(i,j 1,m ,n )-5)&&
(inf_val >= ImLevels(i,j 1,m 1,n )-5)&&
(inf_val >= ImLevels(i,j 1,m-1,n 1)-5)&&
(inf_val >= ImLevels(i,j 1,m ,n 1)-5)&&
(inf_val >= ImLevels(i,j 1,m 1,n 1)-5)
) ) //2、满足26个中极值点*/
if(( (inf_val <= ImLevels(i,j-1,m-1,n-1))&&
(inf_val <= ImLevels(i,j-1,m ,n-1))&&
(inf_val <= ImLevels(i,j-1,m 1,n-1))&&
(inf_val <= ImLevels(i,j-1,m-1,n ))&&
(inf_val <= ImLevels(i,j-1,m ,n ))&&
(inf_val <= ImLevels(i,j-1,m 1,n ))&&
(inf_val <= ImLevels(i,j-1,m-1,n 1))&&
(inf_val <= ImLevels(i,j-1,m ,n 1))&&
(inf_val <= ImLevels(i,j-1,m 1,n 1))&& //底层的小尺度9
(inf_val <= ImLevels(i,j,m-1,n-1))&&
(inf_val <= ImLevels(i,j,m ,n-1))&&
(inf_val <= ImLevels(i,j,m 1,n-1))&&
(inf_val <= ImLevels(i,j,m-1,n ))&&
(inf_val <= ImLevels(i,j,m 1,n ))&&
(inf_val <= ImLevels(i,j,m-1,n 1))&&
(inf_val <= ImLevels(i,j,m ,n 1))&&
(inf_val <= ImLevels(i,j,m 1,n 1))&& //当前层8
(inf_val <= ImLevels(i,j 1,m-1,n-1))&&
(inf_val <= ImLevels(i,j 1,m ,n-1))&&
(inf_val <= ImLevels(i,j 1,m 1,n-1))&&
(inf_val <= ImLevels(i,j 1,m-1,n ))&&
(inf_val <= ImLevels(i,j 1,m ,n ))&&
(inf_val <= ImLevels(i,j 1,m 1,n ))&&
(inf_val <= ImLevels(i,j 1,m-1,n 1))&&
(inf_val <= ImLevels(i,j 1,m ,n 1))&&
(inf_val <= ImLevels(i,j 1,m 1,n 1)) //下一层大尺度9
) ||
( (inf_val >= ImLevels(i,j-1,m-1,n-1))&&
(inf_val >= ImLevels(i,j-1,m ,n-1))&&
(inf_val >= ImLevels(i,j-1,m 1,n-1))&&
(inf_val >= ImLevels(i,j-1,m-1,n ))&&
(inf_val >= ImLevels(i,j-1,m ,n ))&&
(inf_val >= ImLevels(i,j-1,m 1,n ))&&
(inf_val >= ImLevels(i,j-1,m-1,n 1))&&
(inf_val >= ImLevels(i,j-1,m ,n 1))&&
(inf_val >= ImLevels(i,j-1,m 1,n 1))&&
(inf_val >= ImLevels(i,j,m-1,n-1))&&
(inf_val >= ImLevels(i,j,m ,n-1))&&
(inf_val >= ImLevels(i,j,m 1,n-1))&&
(inf_val >= ImLevels(i,j,m-1,n ))&&
(inf_val >= ImLevels(i,j,m 1,n ))&&
(inf_val >= ImLevels(i,j,m-1,n 1))&&
(inf_val >= ImLevels(i,j,m ,n 1))&&
(inf_val >= ImLevels(i,j,m 1,n 1))&&
(inf_val >= ImLevels(i,j 1,m-1,n-1))&&
(inf_val >= ImLevels(i,j 1,m ,n-1))&&
(inf_val >= ImLevels(i,j 1,m 1,n-1))&&
(inf_val >= ImLevels(i,j 1,m-1,n ))&&
(inf_val >= ImLevels(i,j 1,m ,n ))&&
(inf_val >= ImLevels(i,j 1,m 1,n ))&&
(inf_val >= ImLevels(i,j 1,m-1,n 1))&&
(inf_val >= ImLevels(i,j 1,m ,n 1))&&
(inf_val >= ImLevels(i,j 1,m 1,n 1))
) ) //2、满足26个中极值点
{
//存储特征点
//必须具有明显的显著性，即必须大于CONTRAST_THRESHOLD=0.02
if (fabs(ImLevels(i,j,m,n))>= CONTRAST_THRESHOLD 0.05)
{
//最后显著处的特征点必须有足够的曲率比，先计算Hessian矩阵
//Tr_H = Dxx Dyy;
//Det_H = Dxx*Dyy - Dxy^2;
float Dxx,Dyy,Dxy,Tr_H,Det_H,curvature_ratio;
Dxx = ImLevels(i,j,m,n-1) ImLevels(i,j,m,n 1)-2.0*ImLevels(i,j,m,n);
Dyy = ImLevels(i,j,m-1,n) ImLevels(i,j,m 1,n)-2.0*ImLevels(i,j,m,n);
Dxy = ImLevels(i,j,m-1,n-1) ImLevels(i,j,m 1,n 1) - ImLevels(i,j,m 1,n-1) - ImLevels(i,j,m-1,n 1);
Tr_H = Dxx Dyy;
Det_H = Dxx*Dyy - Dxy*Dxy;
//计算主曲率
curvature_ratio=(1.0*Tr_H*Tr_H)/Det_H;
if ((Det_H>=0.0) && (curvature_ratio <= curvature_threshold))//最后得到最具有显著性特征的特征点
{
//将其存储起来，以计算后面的特征描述
keypoint_count ;
Keypoint k;
//为特征点分配内存
k=(Keypoint) malloc(sizeof(struct KeypointSt));
k->next=aa.keypoints1;
aa.keypoints1=k;
k->row=m*(Gaussianpyr1[i].subsample);
k->col=n*(Gaussianpyr1[i].subsample);
k->sy=m;//行
k->sx=n;//列
k->octave=i;
k->level=j;
k->scale=(Gaussianpyr1[i].Octave)[j].absolute_sigma;
}
}
}
}
}
}
}
}
return keypoint_count;
}
int stitchimage::DetectKeypoint2(int numoctaves,ImageOctaves *Gaussianpyr2)
{
aa.keypoints2=NULL;
//计算用于DOG极值点检测的主曲率比的阈值
double curvature_threshold=((CURVATURE_THRESHOLD 1)*(CURVATURE_THRESHOLD 1))/CURVATURE_THRESHOLD;
#define ImLevels(OCTAVE,LEVEL,ROW,COL) ((float *)(DOGoctaves2[(OCTAVE)].Octave[(LEVEL)].Level->data.fl DOGoctaves2[(OCTAVE)].Octave[(LEVEL)].Level->step/sizeof(float) *(ROW)))[(COL)]
int keypoint_count=0;
for (int i=0;i<numoctaves;i )
{
for (int j=1;j<SCALESPEROCTAVE 1;j )//取中间的scaleperoctave个层
{
//在图像的有效区域内寻找具有显著性特征的局部最大值
int dim=(int)(0.5 ((Gaussianpyr2[i].Octave)[j].levelsigmalength) 0.5);
for (int m=dim;m<((DOGoctaves2[i].row)-dim);m )
{
for (int n=dim;n<((DOGoctaves2[i].col)-dim);n )
{
if (ImLevels(i,j,m,n)!=0.0)//first 非零
{
float inf_val=ImLevels(i,j,m,n);
if(( (inf_val <= ImLevels(i,j-1,m-1,n-1))&&
(inf_val <= ImLevels(i,j-1,m ,n-1))&&
(inf_val <= ImLevels(i,j-1,m 1,n-1))&&
(inf_val <= ImLevels(i,j-1,m-1,n ))&&
(inf_val <= ImLevels(i,j-1,m ,n ))&&
(inf_val <= ImLevels(i,j-1,m 1,n ))&&
(inf_val <= ImLevels(i,j-1,m-1,n 1))&&
(inf_val <= ImLevels(i,j-1,m ,n 1))&&
(inf_val <= ImLevels(i,j-1,m 1,n 1))&& //底层的小尺度9
(inf_val <= ImLevels(i,j,m-1,n-1))&&
(inf_val <= ImLevels(i,j,m ,n-1))&&
(inf_val <= ImLevels(i,j,m 1,n-1))&&
(inf_val <= ImLevels(i,j,m-1,n ))&&
(inf_val <= ImLevels(i,j,m 1,n ))&&
(inf_val <= ImLevels(i,j,m-1,n 1))&&
(inf_val <= ImLevels(i,j,m ,n 1))&&
(inf_val <= ImLevels(i,j,m 1,n 1))&& //当前层8
(inf_val <= ImLevels(i,j 1,m-1,n-1))&&
(inf_val <= ImLevels(i,j 1,m ,n-1))&&
(inf_val <= ImLevels(i,j 1,m 1,n-1))&&
(inf_val <= ImLevels(i,j 1,m-1,n ))&&
(inf_val <= ImLevels(i,j 1,m ,n ))&&
(inf_val <= ImLevels(i,j 1,m 1,n ))&&
(inf_val <= ImLevels(i,j 1,m-1,n 1))&&
(inf_val <= ImLevels(i,j 1,m ,n 1))&&
(inf_val <= ImLevels(i,j 1,m 1,n 1)) //下一层大尺度9
) ||
( (inf_val >= ImLevels(i,j-1,m-1,n-1))&&
(inf_val >= ImLevels(i,j-1,m ,n-1))&&
(inf_val >= ImLevels(i,j-1,m 1,n-1))&&
(inf_val >= ImLevels(i,j-1,m-1,n ))&&
(inf_val >= ImLevels(i,j-1,m ,n ))&&
(inf_val >= ImLevels(i,j-1,m 1,n ))&&
(inf_val >= ImLevels(i,j-1,m-1,n 1))&&
(inf_val >= ImLevels(i,j-1,m ,n 1))&&
(inf_val >= ImLevels(i,j-1,m 1,n 1))&&
(inf_val >= ImLevels(i,j,m-1,n-1))&&
(inf_val >= ImLevels(i,j,m ,n-1))&&
(inf_val >= ImLevels(i,j,m 1,n-1))&&
(inf_val >= ImLevels(i,j,m-1,n ))&&
(inf_val >= ImLevels(i,j,m 1,n ))&&
(inf_val >= ImLevels(i,j,m-1,n 1))&&
(inf_val >= ImLevels(i,j,m ,n 1))&&
(inf_val >= ImLevels(i,j,m 1,n 1))&&
(inf_val >= ImLevels(i,j 1,m-1,n-1))&&
(inf_val >= ImLevels(i,j 1,m ,n-1))&&
(inf_val >= ImLevels(i,j 1,m 1,n-1))&&
(inf_val >= ImLevels(i,j 1,m-1,n ))&&
(inf_val >= ImLevels(i,j 1,m ,n ))&&
(inf_val >= ImLevels(i,j 1,m 1,n ))&&
(inf_val >= ImLevels(i,j 1,m-1,n 1))&&
(inf_val >= ImLevels(i,j 1,m ,n 1))&&
(inf_val >= ImLevels(i,j 1,m 1,n 1))
) ) //2、满足26个中极值点
{
//存储特征点
//必须具有明显的显著性，即必须大于CONTRAST_THRESHOLD=0.02
if (fabs(ImLevels(i,j,m,n))>= CONTRAST_THRESHOLD 0.05)
{
//最后显著处的特征点必须有足够的曲率比，先计算Hessian矩阵
//Tr_H = Dxx Dyy;
//Det_H = Dxx*Dyy - Dxy^2;
float Dxx,Dyy,Dxy,Tr_H,Det_H,curvature_ratio;
Dxx = ImLevels(i,j,m,n-1) ImLevels(i,j,m,n 1)-2.0*ImLevels(i,j,m,n);
Dyy = ImLevels(i,j,m-1,n) ImLevels(i,j,m 1,n)-2.0*ImLevels(i,j,m,n);
Dxy = ImLevels(i,j,m-1,n-1) ImLevels(i,j,m 1,n 1) - ImLevels(i,j,m 1,n-1) - ImLevels(i,j,m-1,n 1);
Tr_H = Dxx Dyy;
Det_H = Dxx*Dyy - Dxy*Dxy;
//计算主曲率
curvature_ratio=(1.0*Tr_H*Tr_H)/Det_H;
if ((Det_H>=0.0) && (curvature_ratio <= curvature_threshold))//最后得到最具有显著性特征的特征点
{
//将其存储起来，以计算后面的特征描述
keypoint_count ;
Keypoint k;
//为特征点分配内存
k=(Keypoint) malloc(sizeof(struct KeypointSt));
k->next=aa.keypoints2;
aa.keypoints2=k;
k->row=m*(Gaussianpyr2[i].subsample);
k->col=n*(Gaussianpyr2[i].subsample);
k->sy=m;//行
k->sx=n;//列
k->octave=i;
k->level=j;
k->scale=(Gaussianpyr2[i].Octave)[j].absolute_sigma;
}
}
}
}
}
}
}
}
return keypoint_count;
}
void stitchimage::DisplayKeypointLocation1(IplImage* image, ImageOctaves *Gaussianpyr1)
{
Keypoint p = aa.keypoints1; // p指向第一个结点
while(p) // 没到表尾
{
cvLine( image, cvPoint((int)((p->col)-3),(int)(p->row)),
cvPoint((int)((p->col) 3),(int)(p->row)), CV_RGB(255,255,0),
1, 8, 0 );
cvLine( image, cvPoint((int)(p->col),(int)((p->row)-3)),
cvPoint((int)(p->col),(int)((p->row) 3)), CV_RGB(255,255,0),
1, 8, 0 );
// cvCircle(image,cvPoint((uchar)(p->col),(uchar)(p->row)),
// (int)((GaussianPyr[p->octave].Octave)[p->level].absolute_sigma),
// CV_RGB(255,0,0),1,8,0);
p=p->next;
}
}
void stitchimage::DisplayKeypointLocation2(IplImage* image, ImageOctaves *Gaussianpyr2)
{
Keypoint p = aa.keypoints2; // p指向第一个结点
while(p) // 没到表尾
{
cvLine( image, cvPoint((int)((p->col)-3),(int)(p->row)),
cvPoint((int)((p->col) 3),(int)(p->row)), CV_RGB(255,255,0),
1, 8, 0 );
cvLine( image, cvPoint((int)(p->col),(int)((p->row)-3)),
cvPoint((int)(p->col),(int)((p->row) 3)), CV_RGB(255,255,0),
1, 8, 0 );
// cvCircle(image,cvPoint((uchar)(p->col),(uchar)(p->row)),
// (int)((GaussianPyr[p->octave].Octave)[p->level].absolute_sigma),
// CV_RGB(255,0,0),1,8,0);
p=p->next;
}
}
void stitchimage::Onstep31()
{
// TODO: Add your control notification handler code here
int keycount=DetectKeypoint1(aa.numoctaves,Gaussianpyr1);
cvCopy(imageptocy1,image1,NULL);
DisplayKeypointLocation1(image1,Gaussianpyr1);
cvPyrUp(image1,DoubleSizeImage1,CV_GAUSSIAN_5x5);
//cvReleaseImage(&image1);
cvSaveImage("D:\DoubleSizeImage1.jpg",DoubleSizeImage1);
cvNamedWindow("image1",2);
cvShowImage("image1", DoubleSizeImage1);
cvReleaseImage(&DoubleSizeImage1);
cvWaitKey(0);
cvDestroyWindow("image1");
}
void stitchimage::Onstep32()
{
// TODO: Add your control notification handler code here
int keycount=DetectKeypoint2(aa.numoctaves,Gaussianpyr2);
cvCopy(imageptocy2,image2,NULL);
DisplayKeypointLocation2(image2,Gaussianpyr2);
cvPyrUp(image2,DoubleSizeImage2,CV_GAUSSIAN_5x5);
//cvReleaseImage(&image2);
cvSaveImage("D:\DoubleSizeImage2.jpg",DoubleSizeImage2);
cvNamedWindow("image2",2);
cvShowImage("image2", DoubleSizeImage2);
cvReleaseImage(&DoubleSizeImage2);
cvWaitKey(0);
cvDestroyWindow("image2");
}
void stitchimage::ComputeGrad_DirecandMag1(int numoctaves,ImageOctaves *Gaussianpyr1)
{
mag_pyr1=(ImageOctaves *) malloc(numoctaves * sizeof(ImageOctaves));
grad_pyr1=(ImageOctaves *) malloc(numoctaves * sizeof(ImageOctaves));
#define ImLevels(OCTAVE,LEVEL,ROW,COL) ((float *)(Gaussianpyr1[(OCTAVE)].Octave[(LEVEL)].Level->data.fl Gaussianpyr1[(OCTAVE)].Octave[(LEVEL)].Level->step/sizeof(float) *(ROW)))[(COL)]
for (int i=0;i<numoctaves;i )
{
mag_pyr1[i].Octave=(ImageLevels *) malloc((SCALESPEROCTAVE) * sizeof(ImageLevels));
grad_pyr1[i].Octave=(ImageLevels *) malloc((SCALESPEROCTAVE) * sizeof(ImageLevels));
for (int j=1;j<SCALESPEROCTAVE 1;j )//取中间的SCALESPEROCTAVE个层
{
CvMat *Mag=cvCreateMat(Gaussianpyr1[i].row,Gaussianpyr1[i].col,CV_32FC1);
CvMat *Ori=cvCreateMat(Gaussianpyr1[i].row,Gaussianpyr1[i].col,CV_32FC1);
CvMat *tempMat11=cvCreateMat(Gaussianpyr1[i].row,Gaussianpyr1[i].col,CV_32FC1);
CvMat *tempMat22=cvCreateMat(Gaussianpyr1[i].row,Gaussianpyr1[i].col,CV_32FC1);
cvZero(Mag);
cvZero(Ori);
cvZero(tempMat11);
cvZero(tempMat22);
#define MAG(ROW,COL) ((float *)(Mag->data.fl Mag->step/sizeof(float) *(ROW)))[(COL)]
#define ORI(ROW,COL) ((float *)(Ori->data.fl Ori->step/sizeof(float) *(ROW)))[(COL)]
#define TEMPMAT1(ROW,COL) ((float *)(tempMat11->data.fl tempMat11->step/sizeof(float) *(ROW)))[(COL)]
#define TEMPMAT2(ROW,COL) ((float *)(tempMat22->data.fl tempMat22->step/sizeof(float) *(ROW)))[(COL)]
for (int m=1;m<(Gaussianpyr1[i].row-1);m )
{
for (int n=1;n<(Gaussianpyr1[i].col-1);n )
{
//计算幅值
TEMPMAT1(m,n)=0.5*(ImLevels(i,j,m,n 1)-ImLevels(i,j,m,n-1));//dx
TEMPMAT2(m,n)=0.5*(ImLevels(i,j,m 1,n)-ImLevels(i,j,m-1,n));//dy
MAG(m,n)=sqrt(TEMPMAT1(m,n)*TEMPMAT1(m,n) TEMPMAT2(m,n)*TEMPMAT2(m,n));//幅值
//计算方向
ORI(m,n)=atan(TEMPMAT2(m,n)/TEMPMAT1(m,n));
if (ORI(m,n)==CV_PI)
{
ORI(m,n)=-CV_PI;
}
}
}
((mag_pyr1[i].Octave)[j-1]).Level=Mag;
((grad_pyr1[i].Octave)[j-1]).Level=Ori;
cvReleaseMat(&tempMat11);
cvReleaseMat(&tempMat22);
cvReleaseMat(&Mag);
cvReleaseMat(&Ori);
}
free(mag_pyr1[i].Octave);
free(grad_pyr1[i].Octave);
}
free(mag_pyr1);
free(grad_pyr1);
}
void stitchimage::ComputeGrad_DirecandMag2(int numoctaves,ImageOctaves *Gaussianpyr2)
{
mag_pyr2=(ImageOctaves *) malloc(numoctaves * sizeof(ImageOctaves));
grad_pyr2=(ImageOctaves *) malloc(numoctaves * sizeof(ImageOctaves));
#define ImLevels(OCTAVE,LEVEL,ROW,COL) ((float *)(Gaussianpyr2[(OCTAVE)].Octave[(LEVEL)].Level->data.fl Gaussianpyr2[(OCTAVE)].Octave[(LEVEL)].Level->step/sizeof(float) *(ROW)))[(COL)]
for (int i=0;i<numoctaves;i )
{
mag_pyr2[i].Octave=(ImageLevels *) malloc((SCALESPEROCTAVE) * sizeof(ImageLevels));
grad_pyr2[i].Octave=(ImageLevels *) malloc((SCALESPEROCTAVE) * sizeof(ImageLevels));
for (int j=1;j<SCALESPEROCTAVE 1;j )//取中间的SCALESPEROCTAVE个层
{
CvMat *Mag=cvCreateMat(Gaussianpyr2[i].row,Gaussianpyr2[i].col,CV_32FC1);
CvMat *Ori=cvCreateMat(Gaussianpyr2[i].row,Gaussianpyr2[i].col,CV_32FC1);
CvMat *tempMat11=cvCreateMat(Gaussianpyr2[i].row,Gaussianpyr2[i].col,CV_32FC1);
CvMat *tempMat22=cvCreateMat(Gaussianpyr2[i].row,Gaussianpyr2[i].col,CV_32FC1);
cvZero(Mag);
cvZero(Ori);
cvZero(tempMat11);
cvZero(tempMat22);
#define MAG(ROW,COL) ((float *)(Mag->data.fl Mag->step/sizeof(float) *(ROW)))[(COL)]
#define ORI(ROW,COL) ((float *)(Ori->data.fl Ori->step/sizeof(float) *(ROW)))[(COL)]
#define TEMPMAT1(ROW,COL) ((float *)(tempMat11->data.fl tempMat11->step/sizeof(float) *(ROW)))[(COL)]
#define TEMPMAT2(ROW,COL) ((float *)(tempMat22->data.fl tempMat22->step/sizeof(float) *(ROW)))[(COL)]
for (int m=1;m<(Gaussianpyr2[i].row-1);m )
{
for (int n=1;n<(Gaussianpyr2[i].col-1);n )
{
//计算幅值
TEMPMAT1(m,n)=0.5*(ImLevels(i,j,m,n 1)-ImLevels(i,j,m,n-1));//dx
TEMPMAT2(m,n)=0.5*(ImLevels(i,j,m 1,n)-ImLevels(i,j,m-1,n));//dy
MAG(m,n)=sqrt(TEMPMAT1(m,n)*TEMPMAT1(m,n) TEMPMAT2(m,n)*TEMPMAT2(m,n));//幅值
//计算方向
ORI(m,n)=atan(TEMPMAT2(m,n)/TEMPMAT1(m,n));
if (ORI(m,n)==CV_PI)
{
ORI(m,n)=-CV_PI;
}
}
}
((mag_pyr2[i].Octave)[j-1]).Level=Mag;
((grad_pyr2[i].Octave)[j-1]).Level=Ori;
cvReleaseMat(&tempMat11);
cvReleaseMat(&tempMat22);
cvReleaseMat(&Mag);
cvReleaseMat(&Ori);
}
free(mag_pyr2[i].Octave);
free(grad_pyr2[i].Octave);
}
free(mag_pyr2);
free(grad_pyr2);
}
void stitchimage::AssignTheMainOrientation1(int numoctaves, ImageOctaves *Gaussianpyr,ImageOctaves *mag_pyr,ImageOctaves *grad_pyr)
{
aa.keyDescriptors1=NULL;
int num_bins=36;
float hist_step=2.0*pi/num_bins;
float hist_orient[36];
for (int i=0;i<36;i )
{
hist_orient[i]=-pi i*hist_step;
}
float sigma1=(((Gaussianpyr[0].Octave)[SCALESPEROCTAVE].absolute_sigma))/(Gaussianpyr[0].subsample);//SCALESPEROCTAVE 2
int zero_pad=(int)(max(3.0f, 2 * GAUSSKERN *sigma1 1.0f)*0.5 0.5);
#define ImLevels(OCTAVES,LEVELS,ROW,COL) ((float *)((Gaussianpyr[(OCTAVES)].Octave[(LEVELS)].Level)->data.fl (Gaussianpyr[(OCTAVES)].Octave[(LEVELS)].Level)->step/sizeof(float) *(ROW)))[(COL)]
int keypoint_count=0;
Keypoint p=aa.keypoints1;//p指向第一个结点
while(p)//没到表尾
{
int i=p->octave;
int j=p->level;
int m=p->sy;
int n=p->sx;
if ((m>=zero_pad)&&(m<Gaussianpyr[i].row-zero_pad)&&
(n>=zero_pad)&&(n<Gaussianpyr[i].col-zero_pad))
{
float sigma=( ((Gaussianpyr[i].Octave)[j].absolute_sigma) ) / (Gaussianpyr[i].subsample);
//产生二维高斯模板
CvMat * mat=aa.GaussianKernel2D(sigma);
int dim=(int)(0.5*(mat->rows));
//分配用于存储patch幅值和方向的空间
#define MAT(ROW,COL) ((float *)(mat->data.fl mat->step/sizeof(float) *(ROW)))[(COL)]
//声明方向直方图变量
double * orienthist=(double *) malloc(36*sizeof(double));
for (int sw=0;sw<36; sw)
{
orienthist[sw]=0.0;
}
//在特征点的周围统计梯度方向
for (int x=m-dim,mm=0;x<=(m dim);x ,mm )
{
for (int y=n-dim,nn=0;y<=(n dim);y ,nn )
{
//计算特征点处的幅值
double dx=0.5*(ImLevels(i,j,x,y 1)-ImLevels(i,j,x,y-1));//dx
double dy=0.5*(ImLevels(i,j,x 1,y)-ImLevels(i,j,x-1,y));//dy
double mag=sqrt(dx*dx dy*dy);//幅值
//计算方向
double Ori=atan(1.0*dy/dx);//注意dx!=0
int binIdx=aa.FindClosestRotationBin(36,Ori);
//利用高斯加权累加进直方图相应的块
orienthist[binIdx]=orienthist[binIdx] 1.0*mag*MAT(mm,nn);
}
}
//找出方向直方图中的峰值
aa.AverageWeakBins(orienthist,36);
//找出最大值
double maxGrad=0.0;
int maxBin=0;
for (int b=0;b<36; b)
{
if (orienthist[b] > maxGrad)
{
maxGrad=orienthist[b];
maxBin=b;
}
}
double maxPeakValue=0.0;
double maxDegreeOriention=0.0;
/*if((InterpolateOrientation(orienthist[maxBin==0 ? (36-1) : (maxBin-1)],orienthist[maxBin],
orienthist[(maxBin 1) % 36],&maxDegreeOriention,&maxPeakValue))==false)
printf("BUG: Parabola fitting broken");*/
bool binIsKeypoint[36];
for (b=0;b<36; b)
{
binIsKeypoint[b]=false;
if (b==maxBin)
{
binIsKeypoint[b]=true;
continue;
}
if(orienthist[b] < (peakRelThresh * maxPeakValue))
continue;
int leftI=(b==0) ? (36-1) :(b-1);
int rightI=(b 1) % 36;
if(orienthist[b]<=orienthist[leftI] || orienthist[b] <= orienthist[rightI])
continue;
binIsKeypoint[b]=true;
}
//寻找另一个可能的点
double oneBinRad=(2.0 * pi) / 36;
for (b=0;b<36; b)
{
if (binIsKeypoint[b]==false)
continue;
int bLeft=(b==0) ? (36-1) : (b-1);
int bRight=(b 1) % 36;
double peakValue;
double degreeCorrection;
double maxPeakValue,maxDegreeOriention;
if (aa.InterpolateOrientation ( orienthist[maxBin == 0 ? (36 - 1) : (maxBin - 1)],
orienthist[maxBin], orienthist[(maxBin 1) % 36],
°reeCorrection, &peakValue) == false)
{
printf("BUG: Parabola fitting broken");
}
double degree=(b degreeCorrection) * oneBinRad-pi;
if(degree < -pi)
degree =2.0*pi;
else if(degree > pi)
degree-=2.0*pi;
//存储方向，可以直接利用检测到的链表进行该步主方向的指定；
//分配内存重新存储特征点
Keypoint k;
//allocate memory for the keypoint descriptor
k=(Keypoint) malloc(sizeof(struct KeypointSt));
k->next=aa.keyDescriptors1;
aa.keyDescriptors1=k;
k->descrip=(float*)malloc(LEN * sizeof(float));
k->row=p->row;
k->col=p->col;
k->sy = p->sy; //行
k->sx = p->sx; //列
k->octave = p->octave;
k->level = p->level;
k->scale = p->scale;
k->ori = degree;
k->mag = peakValue;
}
free(orienthist);
}
p=p->next;
}
}
void stitchimage::DisplayOrientation1(IplImage *image,ImageOctaves *Gaussianpyr)
{
p1=aa.keyDescriptors1;//p指向第一个节点,keyDescriptors一定要初始化为零
while (p1)//没到表尾
{
float scale=(Gaussianpyr[p1->octave].Octave)[p1->level].absolute_sigma;
float autoscale=3.0;
float uu=autoscale*scale*cos(p1->ori);
float vv=autoscale*scale*sin(p1->ori);
float x=(p1->col) uu;
float y=(p1->row) vv;
cvLine(image, cvPoint((int)(p1->col),(int)(p1->row)),
cvPoint((int)x,(int)y), CV_RGB(255,255,0),
1, 8, 0 );
float alpha=0.33;
float beta=0.33;
float xx0=(p1->col) uu-alpha*(uu beta*vv);
float yy0=(p1->row) vv-alpha*(vv-beta*uu);
float xx1=(p1->col) uu-alpha*(uu-beta*vv);
float yy1=(p1->row) vv-alpha*(vv beta*uu);
cvLine(image,cvPoint((int)xx0,(int)yy0),cvPoint((int)x,(int)y),CV_RGB(255,255,0),1,8,0);
cvLine(image,cvPoint((int)xx1,(int)yy1),cvPoint((int)x,(int)y),CV_RGB(255,255,0),1,8,0);
p1=p1->next;
}
}
void stitchimage::Onstep41()
{
// TODO: Add your control notification handler code here
ComputeGrad_DirecandMag1(aa.numoctaves,Gaussianpyr1);
AssignTheMainOrientation1(aa.numoctaves,Gaussianpyr1,mag_pyr1,grad_pyr1);
cvCopy(imageptocy1,image1,NULL);
DisplayOrientation1(image1,Gaussianpyr1);
cvSaveImage("D:\mag1.jpg",image1);
cvNamedWindow("image11",2);
cvShowImage("image11",image1);
//cvReleaseImage(& image1);
cvWaitKey(0);
cvDestroyWindow("image11");
}
void stitchimage::AssignTheMainOrientation2(int numoctaves, ImageOctaves *Gaussianpyr,ImageOctaves *mag_pyr,ImageOctaves *grad_pyr)
{
aa.keyDescriptors2=NULL;
int num_bins=36;
float hist_step=2.0*pi/num_bins;
float hist_orient[36];
for (int i=0;i<36;i )
{
hist_orient[i]=-pi i*hist_step;
}
float sigma1=(((Gaussianpyr[0].Octave)[SCALESPEROCTAVE].absolute_sigma))/(Gaussianpyr[0].subsample);//SCALESPEROCTAVE 2
int zero_pad=(int)(max(3.0f, 2 * GAUSSKERN *sigma1 1.0f)*0.5 0.5);
#define ImLevels(OCTAVES,LEVELS,ROW,COL) ((float *)((Gaussianpyr[(OCTAVES)].Octave[(LEVELS)].Level)->data.fl (Gaussianpyr[(OCTAVES)].Octave[(LEVELS)].Level)->step/sizeof(float) *(ROW)))[(COL)]
int keypoint_count=0;
Keypoint p=aa.keypoints2;//p指向第一个结点
while(p)//没到表尾
{
int i=p->octave;
int j=p->level;
int m=p->sy;
int n=p->sx;
if ((m>=zero_pad)&&(m<Gaussianpyr[i].row-zero_pad)&&
(n>=zero_pad)&&(n<Gaussianpyr[i].col-zero_pad))
{
float sigma=( ((Gaussianpyr[i].Octave)[j].absolute_sigma) ) / (Gaussianpyr[i].subsample);
//产生二维高斯模板
CvMat * mat=aa.GaussianKernel2D(sigma);
int dim=(int)(0.5*(mat->rows));
//分配用于存储patch幅值和方向的空间
#define MAT(ROW,COL) ((float *)(mat->data.fl mat->step/sizeof(float) *(ROW)))[(COL)]
//声明方向直方图变量
double * orienthist=(double *) malloc(36*sizeof(double));
for (int sw=0;sw<36; sw)
{
orienthist[sw]=0.0;
}
//在特征点的周围统计梯度方向
for (int x=m-dim,mm=0;x<=(m dim);x ,mm )
{
for (int y=n-dim,nn=0;y<=(n dim);y ,nn )
{
//计算特征点处的幅值
double dx=0.5*(ImLevels(i,j,x,y 1)-ImLevels(i,j,x,y-1));//dx
double dy=0.5*(ImLevels(i,j,x 1,y)-ImLevels(i,j,x-1,y));//dy
double mag=sqrt(dx*dx dy*dy);//幅值
//计算方向
double Ori=atan(1.0*dy/dx);//注意dx!=0
int binIdx=aa.FindClosestRotationBin(36,Ori);
//利用高斯加权累加进直方图相应的块
orienthist[binIdx]=orienthist[binIdx] 1.0*mag*MAT(mm,nn);
}
}
//找出方向直方图中的峰值
aa.AverageWeakBins(orienthist,36);
//找出最大值
double maxGrad=0.0;
int maxBin=0;
for (int b=0;b<36; b)
{
if (orienthist[b] > maxGrad)
{
maxGrad=orienthist[b];
maxBin=b;
}
}
double maxPeakValue=0.0;
double maxDegreeOriention=0.0;
/*if((InterpolateOrientation(orienthist[maxBin==0 ? (36-1) : (maxBin-1)],orienthist[maxBin],
orienthist[(maxBin 1) % 36],&maxDegreeOriention,&maxPeakValue))==false)
printf("BUG: Parabola fitting broken");*/
bool binIsKeypoint[36];
for (b=0;b<36; b)
{
binIsKeypoint[b]=false;
if (b==maxBin)
{
binIsKeypoint[b]=true;
continue;
}
if(orienthist[b] < (peakRelThresh * maxPeakValue))
continue;
int leftI=(b==0) ? (36-1) :(b-1);
int rightI=(b 1) % 36;
if(orienthist[b]<=orienthist[leftI] || orienthist[b] <= orienthist[rightI])
continue;
binIsKeypoint[b]=true;
}
//寻找另一个可能的点
double oneBinRad=(2.0 * pi) / 36;
for (b=0;b<36; b)
{
if (binIsKeypoint[b]==false)
continue;
int bLeft=(b==0) ? (36-1) : (b-1);
int bRight=(b 1) % 36;
double peakValue;
double degreeCorrection;
double maxPeakValue,maxDegreeOriention;
if (aa.InterpolateOrientation ( orienthist[maxBin == 0 ? (36 - 1) : (maxBin - 1)],
orienthist[maxBin], orienthist[(maxBin 1) % 36],
°reeCorrection, &peakValue) == false)
{
printf("BUG: Parabola fitting broken");
}
double degree=(b degreeCorrection) * oneBinRad-pi;
if(degree < -pi)
degree =2.0*pi;
else if(degree > pi)
degree-=2.0*pi;
//存储方向，可以直接利用检测到的链表进行该步主方向的指定；
//分配内存重新存储特征点
Keypoint k;
//allocate memory for the keypoint descriptor
k=(Keypoint) malloc(sizeof(struct KeypointSt));
k->next=aa.keyDescriptors2;
aa.keyDescriptors2=k;
k->descrip=(float*)malloc(LEN * sizeof(float));
k->row=p->row;
k->col=p->col;
k->sy = p->sy; //行
k->sx = p->sx; //列
k->octave = p->octave;
k->level = p->level;
k->scale = p->scale;
k->ori = degree;
k->mag = peakValue;
}
free(orienthist);
}
p=p->next;
}
}
void stitchimage::DisplayOrientation2(IplImage *image,ImageOctaves *Gaussianpyr)
{
p2=aa.keyDescriptors2;//p指向第一个节点,keyDescriptors一定要初始化为零
while (p2)//没到表尾
{
float scale=(Gaussianpyr[p2->octave].Octave)[p2->level].absolute_sigma;
float autoscale=3.0;
float uu=autoscale*scale*cos(p2->ori);
float vv=autoscale*scale*sin(p2->ori);
float x=(p2->col) uu;
float y=(p2->row) vv;
cvLine(image, cvPoint((int)(p2->col),(int)(p2->row)),
cvPoint((int)x,(int)y), CV_RGB(255,255,0),
1, 8, 0 );
float alpha=0.33;
float beta=0.33;
float xx0=(p2->col) uu-alpha*(uu beta*vv);
float yy0=(p2->row) vv-alpha*(vv-beta*uu);
float xx1=(p2->col) uu-alpha*(uu-beta*vv);
float yy1=(p2->row) vv-alpha*(vv beta*uu);
cvLine(image,cvPoint((int)xx0,(int)yy0),cvPoint((int)x,(int)y),CV_RGB(255,255,0),1,8,0);
cvLine(image,cvPoint((int)xx1,(int)yy1),cvPoint((int)x,(int)y),CV_RGB(255,255,0),1,8,0);
p2=p2->next;
}
}
void stitchimage::Onstep42()
{
// TODO: Add your control notification handler code here
ComputeGrad_DirecandMag2(aa.numoctaves,Gaussianpyr2);
AssignTheMainOrientation2(aa.numoctaves,Gaussianpyr2,mag_pyr2,grad_pyr2);
//cvCopy(imageptocy2,image2,NULL);
DisplayOrientation2(image2,Gaussianpyr2);
cvSaveImage("D:\mag2.jpg",image2);
cvNamedWindow("image22",2);//cvNamedWindow("image22",1)1是代表窗口的大小
cvShowImage("image22",image2);
//cvReleaseImage(& image2);
cvWaitKey(0);
cvDestroyWindow("image22");
}
void stitchimage::ExtractFeatureDescriptors1(int numoctaves,ImageOctaves *Gaussianpyr)
{
//主要有八个方向
float orient_bin_spacing=pi/4;
float orient_angles[8]={-pi,-pi orient_bin_spacing,-pi*0.5,-orient_bin_spacing,
0.0,orient_bin_spacing,pi*0.5,pi orient_bin_spacing};
//产生描述子中心各点坐标
float *feat_grid=(float *) malloc(2*16*sizeof(float));
for (int i=0;i<GridSpacing;i )//GridSpacing=4
{
for (int j=0;j<2*GridSpacing; j, j)
{
feat_grid[i*2*GridSpacing j]=-6.0 i*GridSpacing;
feat_grid[i*2*GridSpacing j 1]=-6.0 0.5*j*GridSpacing;
}
}
//产生网格
float *feat_samples=(float *) malloc(2*256*sizeof(float));
for (i=0;i<4*GridSpacing;i )
{
for (int j=0;j<8*GridSpacing;j =2)
{
feat_samples[i*8*GridSpacing j]=-(2*GridSpacing-0.5) i;
feat_samples[i*8*GridSpacing j 1]=-(2*GridSpacing-0.5) 0.5*j;
}
}
float feat_window=2*GridSpacing;
p1=aa.keyDescriptors1;//p1指向第一个节点
while(p1)//没到表尾
{
float scale=(Gaussianpyr[p1->octave].Octave)[p1->level].absolute_sigma;
float sine=sin(p1->ori);
float cosine=cos(p1->ori);
//计算中心点坐标旋转后的位置
float *featcenter=(float *) malloc(2*16*sizeof(float));
for (int i=0;i<GridSpacing;i )
{
for (int j=0;j<2*GridSpacing;j =2)
{
float x=feat_grid[i*2*GridSpacing j];
float y=feat_grid[i*2*GridSpacing j 1];
featcenter[i*2*GridSpacing j]=((cosine * x sine * y) p1->sx);
featcenter[i*2*GridSpacing j 1]=((-sine * x cosine * y) p1->sy);
}
}
float *feat=(float *) malloc(2*256*sizeof(float));
//原程序有误
for (i=0;i<2*64*GridSpacing;i ,i )
{
float x=feat_samples[i];
float y=feat_samples[i 1];
feat[i]=((cosine * x sine * y ) p1->sx);
feat[i 1]=((-sine * x cosine * y) p1->sy);
}
//初始化特征点描述
float *feat_desc=(float *) malloc(128 * sizeof(float));
for (i=0;i<128;i )
{
feat_desc[i]=0.0;
}
for (i=0;i<512; i, i)
{
float x_sample=feat[i];
float y_sample=feat[i 1];
float sample12=aa.getPixelBI(((Gaussianpyr[p1->octave].Octave)[p1->level]).Level, x_sample, y_sample-1);
float sample21=aa.getPixelBI(((Gaussianpyr[p1->octave].Octave)[p1->level]).Level, x_sample-1, y_sample);
float sample22=aa.getPixelBI(((Gaussianpyr[p1->octave].Octave)[p1->level]).Level, x_sample, y_sample);
float sample23=aa.getPixelBI(((Gaussianpyr[p1->octave].Octave)[p1->level]).Level, x_sample 1, y_sample);
float sample32=aa.getPixelBI(((Gaussianpyr[p1->octave].Octave)[p1->level]).Level, x_sample, y_sample 1);
float diff_x=sample23-sample21;
float diff_y=sample32-sample12;
float mag_sample=sqrt(diff_x*diff_x diff_y*diff_y);
float grad_sample=atan(diff_y / diff_x);
if(grad_sample==CV_PI)
grad_sample=-CV_PI;
//计算x，y方向的权重
float *x_wght=(float *) malloc(GridSpacing * GridSpacing * sizeof(float));
float *y_wght=(float *) malloc(GridSpacing * GridSpacing * sizeof(float));
float *pos_wght=(float *) malloc(8*GridSpacing*GridSpacing*sizeof(float));
for (int m=0;m<32; m, m)
{
float x=featcenter[m];
float y=featcenter[m 1];
x_wght[m/2]=max(1-(fabs(x-x_sample)*1.0/GridSpacing),0);
y_wght[m/2]=max(1-(fabs(y-y_sample)*1.0/GridSpacing),0);
}
for ( m=0;m<16; m)
for(int n=0;n<8; n)
pos_wght[m*8 n]=x_wght[m]*y_wght[m];
free(x_wght);
free(y_wght);
//计算方向的加权，首先旋转梯度场到主方向，然后计算差异
float diff[8],orient_wght[128];
for (m=0;m<8; m)
{
float angle=grad_sample-(p1->ori)-orient_angles[m] CV_PI;
float temp=angle / (2.0 * CV_PI);
angle-=(int)(temp) * (2.0*CV_PI);
diff[m]=angle-CV_PI;
}
//计算高斯权重
float x=p1->sx;
float y=p1->sy;
float g=exp(-((x_sample-x)*(x_sample-x) (y-y_sample)*(y-y_sample))/(2*feat_window*feat_window)
)/(2*CV_PI*feat_window*feat_window);
for (m=0;m<128; m)
{
orient_wght[m]=max((1.0 - 1.0*fabs(diff[m%8])/orient_bin_spacing),0);
feat_desc[m]=feat_desc[m] orient_wght[m]*pos_wght[m]*g*mag_sample;
}
free(pos_wght);
}
free(feat);
free(featcenter);
float norm=aa.GetVecNorm(feat_desc,128);
for (int m=0;m<128;m )
{
feat_desc[m]/=norm;
if (feat_desc[m]>0.2)
{
feat_desc[m]=0.2;
}
}
norm=aa.GetVecNorm(feat_desc,128);
for (m=0;m<128;m )
{
feat_desc[m]/=norm;
}
p1->descrip=feat_desc;
p1=p1->next;
//free(feat_desc);//不能使用，否则内存会在同一个地址使用
}
free(feat_grid);
free(feat_samples);
}
void stitchimage::Onstep51()
{
// TODO: Add your control notification handler code here
ExtractFeatureDescriptors1(aa.numoctaves,Gaussianpyr1);
}
void stitchimage::ExtractFeatureDescriptors2(int numoctaves,ImageOctaves *Gaussianpyr)
{
//主要有八个方向
float orient_bin_spacing=pi/4;
float orient_angles[8]={-pi,-pi orient_bin_spacing,-pi*0.5,-orient_bin_spacing,
0.0,orient_bin_spacing,pi*0.5,pi orient_bin_spacing};
//产生描述子中心各点坐标
float *feat_grid=(float *) malloc(2*16*sizeof(float));
for (int i=0;i<GridSpacing;i )//GridSpacing=4
{
for (int j=0;j<2*GridSpacing; j, j)
{
feat_grid[i*2*GridSpacing j]=-6.0 i*GridSpacing;
feat_grid[i*2*GridSpacing j 1]=-6.0 0.5*j*GridSpacing;
}
}
//产生网格
float *feat_samples=(float *) malloc(2*256*sizeof(float));
for (i=0;i<4*GridSpacing;i )
{
for (int j=0;j<8*GridSpacing;j =2)
{
feat_samples[i*8*GridSpacing j]=-(2*GridSpacing-0.5) i;
feat_samples[i*8*GridSpacing j 1]=-(2*GridSpacing-0.5) 0.5*j;
}
}
float feat_window=2*GridSpacing;
p2=aa.keyDescriptors2;//p指向第一个节点
while(p2)//没到表尾
{
float scale=(Gaussianpyr[p2->octave].Octave)[p2->level].absolute_sigma;
float sine=sin(p2->ori);
float cosine=cos(p2->ori);
//计算中心点坐标旋转后的位置
float *featcenter=(float *) malloc(2*16*sizeof(float));
for (int i=0;i<GridSpacing;i )
{
for (int j=0;j<2*GridSpacing;j =2)
{
float x=feat_grid[i*2*GridSpacing j];
float y=feat_grid[i*2*GridSpacing j 1];
featcenter[i*2*GridSpacing j]=((cosine * x sine * y) p2->sx);
featcenter[i*2*GridSpacing j 1]=((-sine * x cosine * y) p2->sy);
}
}
float *feat=(float *) malloc(2*256*sizeof(float));
//原程序有误
for (i=0;i<2*64*GridSpacing;i ,i )
{
float x=feat_samples[i];
float y=feat_samples[i 1];
feat[i]=((cosine * x sine * y ) p2->sx);
feat[i 1]=((-sine * x cosine * y) p2->sy);
}
//初始化特征点描述
//float *feat_desc=(float *) malloc(128 * sizeof(float));
float *feat_desc=(float *) malloc(128 * sizeof(float));
for (i=0;i<128;i )
{
feat_desc[i]=0.0;
}
for (i=0;i<512; i, i)
{
float x_sample=feat[i];
float y_sample=feat[i 1];
float sample12=aa.getPixelBI(((Gaussianpyr[p2->octave].Octave)[p2->level]).Level, x_sample, y_sample-1);
float sample21=aa.getPixelBI(((Gaussianpyr[p2->octave].Octave)[p2->level]).Level, x_sample-1, y_sample);
float sample22=aa.getPixelBI(((Gaussianpyr[p2->octave].Octave)[p2->level]).Level, x_sample, y_sample);
float sample23=aa.getPixelBI(((Gaussianpyr[p2->octave].Octave)[p2->level]).Level, x_sample 1, y_sample);
float sample32=aa.getPixelBI(((Gaussianpyr[p2->octave].Octave)[p2->level]).Level, x_sample, y_sample 1);
float diff_x=sample23-sample21;
float diff_y=sample32-sample12;
float mag_sample=sqrt(diff_x*diff_x diff_y*diff_y);
float grad_sample=atan(diff_y / diff_x);
if(grad_sample==CV_PI)
grad_sample=-CV_PI;
//计算x，y方向的权重
float *x_wght=(float *) malloc(GridSpacing * GridSpacing * sizeof(float));
float *y_wght=(float *) malloc(GridSpacing * GridSpacing * sizeof(float));
float *pos_wght=(float *) malloc(8*GridSpacing*GridSpacing*sizeof(float));
for (int m=0;m<32; m, m)
{
float x=featcenter[m];
float y=featcenter[m 1];
x_wght[m/2]=max(1-(fabs(x-x_sample)*1.0/GridSpacing),0);
y_wght[m/2]=max(1-(fabs(y-y_sample)*1.0/GridSpacing),0);
}
for ( m=0;m<16; m)
for(int n=0;n<8; n)
pos_wght[m*8 n]=x_wght[m]*y_wght[m];
free(x_wght);
free(y_wght);
//计算方向的加权，首先旋转梯度场到主方向，然后计算差异
float diff[8],orient_wght[128];
for (m=0;m<8; m)
{
float angle=grad_sample-(p2->ori)-orient_angles[m] CV_PI;
float temp=angle / (2.0 * CV_PI);
angle-=(int)(temp) * (2.0*CV_PI);
diff[m]=angle-CV_PI;
}
//计算高斯权重
float x=p2->sx;
float y=p2->sy;
float g=exp(-((x_sample-x)*(x_sample-x) (y-y_sample)*(y-y_sample))/(2*feat_window*feat_window)
)/(2*CV_PI*feat_window*feat_window);
for (m=0;m<128; m)
{
orient_wght[m]=max((1.0 - 1.0*fabs(diff[m%8])/orient_bin_spacing),0);
feat_desc[m]=feat_desc[m] orient_wght[m]*pos_wght[m]*g*mag_sample;
}
free(pos_wght);
}
free(feat);
free(featcenter);
float norm=aa.GetVecNorm(feat_desc,128);
for (int m=0;m<128;m )
{
feat_desc[m]/=norm;
if (feat_desc[m]>0.2)
{
feat_desc[m]=0.2;
}
}
norm=aa.GetVecNorm(feat_desc,128);
for (m=0;m<128;m )
{
feat_desc[m]/=norm;
}
p2->descrip=feat_desc;
p2=p2->next;
//free(feat_desc);
}
free(feat_grid);
free(feat_samples);
}
void stitchimage::Onstep52()
{
// TODO: Add your control notification handler code here
ExtractFeatureDescriptors2(aa.numoctaves,Gaussianpyr2);
}
void stitchimage::stitch()
{
p1=aa.keyDescriptors1;
p2=aa.keyDescriptors2;
Keyaxis xy[200];
float dis;
float distance1,distance2;
int k=0;
int x1,y1,x2,y2;
int nn;
int count=0;
int m=0;
float min,min1;
while(p2)
{
k=k 1;
p2=p2->next;
}
while(p1)
{
min=100,min1=100;
float *fest_desc1=(float *) malloc(128 * sizeof(float));
//float *fest_desc1;
//fest_desc1=new float[128];
//memcpy(fest_desc1,p1->descrip,128 * sizeof(float));
/* for (int i=0;i<128;i )
{
fest_desc1[i]=*(p1->descrip);
(p1->descrip) ;
}*/
fest_desc1=p1->descrip;
float sum[200];
for (int j=0;j<200;j )
sum[j]=0.0;
int n=0;
p2=aa.keyDescriptors2;
while(p2)
{
float *fest_desc2=(float *) malloc(128*sizeof(float));
//memcpy(fest_desc2,p2->descrip,128 * sizeof(float));
fest_desc2=p2->descrip;
for (int i=0;i<128;i )
{
dis=fabs(fest_desc1[i]-fest_desc2[i]);
sum[n]=sum[n] dis;
}
if (sum[n]<min)
{
min=sum[n];
distance1=min;
nn=n;
}
n=n 1;
//free(fest_desc2);
p2=p2->next;
}
for (int i=0;i<k;i )
{
if (i!=nn)
{
if (sum[i]<min1)
{
min1=sum[i];
distance2=min1;
}
}
}
if (distance1*1.0/distance2<0.6)
{
n=0;
p2=aa.keyDescriptors2;
while(p2)
{
if (n==nn)
{
xy[m].width1=p1->col;
xy[m].height1=p1->row;
xy[m].width2=p2->col;
xy[m].height2=p2->row;
m ;
count ;
}
n ;
p2=p2->next;
}
}
/*int k1,k2;//冒泡排序
float temp=0;
for (k1=1;k1<k;k1 )
{
for (k2=k-1;k2>=k1;k2--)
{
if (sum[k2]<sum[k2-1])
{
temp=sum[k2-1];
sum[k2-1]=sum[k2];
sum[k2]=temp;
}
}
}*/
//delete fest_desc1;
//free(fest_desc1);
//fest_desc1=NULL;
p1=p1->next;
}
// free(fest_desc1);
int mm=0;
for (int i=0;i<count-1;i )
{
if ((xy[i].width1-xy[i].width2)<100 || (xy[i].width1==xy[i 1].width1))
continue;
xy1[mm].width1=xy[i].width1;
xy1[mm].height1=xy[i].height1;
xy1[mm].width2=xy[i].width2;
xy1[mm].height2=xy[i].height2;
mm ;
}
ncount=mm;
// opencvsolve();
// gauss();
Keyaxis xx;
xx.width1=xy1[0].width1;
xx.height1=xy1[0].height1;
xx.width2=xy1[0].width2;
xx.height2=xy1[0].height2;
stitchok(xx);
}
void stitchimage::stitch1(IplImage *imag1,IplImage* imag2)
{
p1=aa.keyDescriptors1;
p2=aa.keyDescriptors2;
Keyaxis xy[200];
Keypoint p3=NULL;
float dis;
float distance1,distance11,distance2,distance22;
int k,kk;
k=0;
kk=0;
int x1,y1,x2,y2;
int n,nn,nnn,n1;
int count=0;
int m=0;
float min,min1;
while(p2)
{
k=k 1;
p2=p2->next;
}
while(p1)
{
kk=kk 1;
p1=p1->next;
}
p1=aa.keyDescriptors1;
while(p1)
{
p3=p1;
min=100,min1=100;
float *fest_desc1=(float *) malloc(128 * sizeof(float));
fest_desc1=p1->descrip;
float sum[200];
for (int j=0;j<200;j )
sum[j]=0.0;
n=0;
p2=aa.keyDescriptors2;
while(p2)
{
float *fest_desc2=(float *) malloc(128*sizeof(float));
fest_desc2=p2->descrip;
for (int i=0;i<128;i )
{
dis=fabs(fest_desc1[i]-fest_desc2[i]);
sum[n]=sum[n] dis;
}
if (sum[n]<min)
{
min=sum[n];
distance1=min;
nn=n;
}
n=n 1;
p2=p2->next;
}
for (int i=0;i<k;i )
{
if (i!=nn)
{
if (sum[i]<min1)
{
min1=sum[i];
distance2=min1;
}
}
}
n1=0;
p2=aa.keyDescriptors2;
while(p2)
{
if (n1==nn)
{
min=100,min1=100;
float *fest_desc2=(float *) malloc(128 * sizeof(float));
fest_desc2=p2->descrip;
// float sum[200];
for (int j=0;j<200;j )
sum[j]=0.0;
p1=aa.keyDescriptors1;
n=0;
while(p1)
{
float *fest_desc1=(float *) malloc(128*sizeof(float));
fest_desc1=p1->descrip;
for (int i=0;i<128;i )
{
dis=fabs(fest_desc2[i]-fest_desc1[i]);
sum[n]=sum[n] dis;
}
if (sum[n]<min)
{
min=sum[n];
distance11=min;
nnn=n;
}
n=n 1;
p1=p1->next;
}
for (int i=0;i<kk;i )
{
if (i!=nnn)
{
if (sum[i]<min1)
{
min1=sum[i];
distance22=min1;
}
}
}
}
n1 ;
p2=p2->next;
}
p1=p3;
//if ((distance1*1.0/distance2<0.4) && (distance11*1.0/distance22<0.4) && fabs(distance1-distance11)<0.000001)
if ((distance1*1.0/distance2<0.5) && (distance11*1.0/distance22<0.5) && (distance1==distance11))
{
n=0;
p2=aa.keyDescriptors2;
while(p2)
{
if (n==nn)
{
xy[m].width1=p1->col;
xy[m].height1=p1->row;
xy[m].width2=p2->col;
xy[m].height2=p2->row;
m ;
count ;
}
n ;
p2=p2->next;
}
}
p1=p1->next;
}
if (m==0)
{
MessageBox("找不到特征点对");
}
else
{
int wd,hd;
wd=imag1->width-1550;
hd=(imag1->height)/2-600;
Keyaxis xx;
xx.width1=xy[0].width1 wd;
xx.height1=xy[0].height1 hd;
xx.width2=xy[0].width2;
xx.height2=xy[0].height2 hd;
stitchok(xx,imag1,imag2);
}
}
void stitchimage:: stitchok(Keyaxis xx,IplImage* imag1,IplImage* imag2)
{
int width1;
width1=imag1->width;
int x,x1,y,y1,x3,x4,x5,disy;
x=xx.width1;
x1=xx.width2;
x3=width1-x x1;
x4=width1-x3;
x5=width1-x4;
y=xx.height1;
y1=xx.height2;
int dis;
dis=y-y1;
disy=abs(y1-y);
// disy=27;
uchar* data3;
data3=(uchar*)imag1->imageData;
uchar* data4;
data4=(uchar*)imag2->imageData;
int width33,height33;
image33=NULL;
CvSize image33_cvsize;
height33=imag1->height-disy;
width33=imag1->width imag2->width-x3;
image33_cvsize.height=height33;
image33_cvsize.width=width33;
uchar *dataout;
image33 = cvCreateImage(image33_cvsize, imag1->depth,imag1->nChannels);
dataout=(uchar*)image33->imageData;
float d,g;
if (dis<0)
{
for(int i=0;i<height33;i )
{
d=1.0;
for(int j=0;j<x4;j )
{
for(int n=0;n<3;n )
dataout[i*image33->widthStep j*image33->nChannels n]=
data3[i*imag1->widthStep j*imag1->nChannels n];
}
for (j=x4;j<width1;j )
{
g=d-1.0/x3*(j-x4);
for(int n=0;n<3;n )
{
dataout[i*image33->widthStep j*image33->nChannels n]=(int)
(g*data3[i*imag1->widthStep j*imag1->nChannels n]
(1.0-g)*data4[(i disy)*imag2->widthStep (j-x4)*imag2->nChannels n]);
}
}
for (j=width1;j<width33;j )
{
for(int n=0;n<3;n )
dataout[i*image33->widthStep j*image33->nChannels n]=
data4[(i disy)*imag2->widthStep (j-x4)*imag2->nChannels n];
}
}
}
else
{
for(int i=disy;i<height33 disy;i )
{
d=1.0;
for(int j=0;j<x4;j )
{
for(int n=0;n<3;n )
dataout[(i-disy)*image33->widthStep j*image33->nChannels n]=
data3[i*imag1->widthStep j*imag1->nChannels n];
}
for (j=x4;j<width1;j )
{
g=d-1.0/x3*(j-x4);
for(int n=0;n<3;n )
{
dataout[(i-disy)*image33->widthStep j*image33->nChannels n]=(int)
(g*data3[i*imag1->widthStep j*imag1->nChannels n]
(1.0-g)*data4[(i-disy)*imag2->widthStep (j-x4)*imag2->nChannels n]);
}
}
for (j=width1;j<width33;j )
{
for(int n=0;n<3;n )
dataout[(i-disy)*image33->widthStep j*image33->nChannels n]=
data4[(i-disy)*imag2->widthStep (j-x4)*imag2->nChannels n];
}
}
}
/* IplImage* dst1=NULL;
dst1=cvCreateImage(cvGetSize(image33),image33->depth,image33->nChannels);
cvCopy(image33,dst1,NULL);
cvSaveImage("D:/fullimag.jpg",image33);
float wscale = 0.25;
float hscale = 0.22;
CvSize dst1_cvsize;
dst1_cvsize.width = image33->width * wscale; //目标图像的宽为源图象宽的scale倍
dst1_cvsize.height = image33->height * hscale; //目标图像的高为源图象高的scale倍
dst1 = cvCreateImage( dst1_cvsize, image33->depth, image33->nChannels); //构造目标图象
cvResize(image33, dst1, CV_INTER_LINEAR); //缩放源图像到目标图像
cvNamedWindow( "全景图", CV_WINDOW_AUTOSIZE ); //创建用于显示目标图像的窗口
cvShowImage( "全景图", dst1 ); //显示目标图像
cvMoveWindow( "全景图",0,500);
cvWaitKey(-1);
cvReleaseImage(&dst1); //释放目标图像占用的内存
cvDestroyWindow( "全景图" ); //销毁窗口“dst”*/
}
void stitchimage::stitch1()
{
p1=aa.keyDescriptors1;
p2=aa.keyDescriptors2;
Keyaxis xy[200];
Keypoint p3=NULL;
float dis;
float distance1,distance11,distance2,distance22;
int k,kk;
k=0;
kk=0;
int x1,y1,x2,y2;
int n,nn,nnn,n1;
int count=0;
int m=0;
float min,min1;
while(p2)
{
k=k 1;
p2=p2->next;
}
while(p1)
{
kk=kk 1;
p1=p1->next;
}
p1=aa.keyDescriptors1;
while(p1)
{
p3=p1;
min=100,min1=100;
float *fest_desc1=(float *) malloc(128 * sizeof(float));
fest_desc1=p1->descrip;
float sum[200];
for (int j=0;j<200;j )
sum[j]=0.0;
n=0;
p2=aa.keyDescriptors2;
while(p2)
{
float *fest_desc2=(float *) malloc(128*sizeof(float));
fest_desc2=p2->descrip;
for (int i=0;i<128;i )
{
dis=fabs(fest_desc1[i]-fest_desc2[i]);
sum[n]=sum[n] dis;
}
if (sum[n]<min)
{
min=sum[n];
distance1=min;
nn=n;
}
n=n 1;
p2=p2->next;
}
for (int i=0;i<k;i )
{
if (i!=nn)
{
if (sum[i]<min1)
{
min1=sum[i];
distance2=min1;
}
}
}
n1=0;
p2=aa.keyDescriptors2;
while(p2)
{
if (n1==nn)
{
min=100,min1=100;
float *fest_desc2=(float *) malloc(128 * sizeof(float));
fest_desc2=p2->descrip;
// float sum[200];
for (int j=0;j<200;j )
sum[j]=0.0;
p1=aa.keyDescriptors1;
n=0;
while(p1)
{
float *fest_desc1=(float *) malloc(128*sizeof(float));
fest_desc1=p1->descrip;
for (int i=0;i<128;i )
{
dis=fabs(fest_desc2[i]-fest_desc1[i]);
sum[n]=sum[n] dis;
}
if (sum[n]<min)
{
min=sum[n];
distance11=min;
nnn=n;
}
n=n 1;
p1=p1->next;
}
for (int i=0;i<kk;i )
{
if (i!=nnn)
{
if (sum[i]<min1)
{
min1=sum[i];
distance22=min1;
}
}
}
}
n1 ;
p2=p2->next;
}
p1=p3;
//if ((distance1*1.0/distance2<0.4) && (distance11*1.0/distance22<0.4) && fabs(distance1-distance11)<0.000001)
if ((distance1*1.0/distance2<0.4) && (distance11*1.0/distance22<0.4) && (distance1==distance11))
{
n=0;
p2=aa.keyDescriptors2;
while(p2)
{
if (n==nn)
{
xy[m].width1=p1->col;
xy[m].height1=p1->row;
xy[m].width2=p2->col;
xy[m].height2=p2->row;
m ;
count ;
}
n ;
p2=p2->next;
}
}
p1=p1->next;
}
if (m==0)
{
MessageBox("找不到特征点对");
}
else
{
int wd,hd;
wd=imageptocy11->width-1550;
hd=(imageptocy11->height)/2-600;
Keyaxis xx;
xx.width1=xy[0].width1 wd;
xx.height1=xy[0].height1 hd;
xx.width2=xy[0].width2;
xx.height2=xy[0].height2 hd;
stitchok(xx);
}
}
void stitchimage::stitch2()
{
p1=aa.keyDescriptors1;
p2=aa.keyDescriptors2;
Keyaxis xy[200];
Keypoint p3=NULL;
float dis;
float distance1,distance11,distance2,distance22;
int k,kk;
k=0;
kk=0;
int x1,y1,x2,y2;
int n,nn,nnn,n1;
int count=0;
int m=0;
float min,min1;
while(p2)
{
k=k 1;
p2=p2->next;
}
while(p1)
{
kk=kk 1;
p1=p1->next;
}
p1=aa.keyDescriptors1;
while(p1)
{
p3=p1;
min=100,min1=100;
float *fest_desc1=(float *) malloc(128 * sizeof(float));
fest_desc1=p1->descrip;
float sum[200];
for (int j=0;j<200;j )
sum[j]=0.0;
n=0;
p2=aa.keyDescriptors2;
while(p2)
{
float *fest_desc2=(float *) malloc(128*sizeof(float));
fest_desc2=p2->descrip;
for (int i=0;i<128;i )
{
dis=fabs(fest_desc1[i]-fest_desc2[i]);
sum[n]=sum[n] dis;
}
if (sum[n]<min)
{
min=sum[n];
distance1=min;
nn=n;
}
n=n 1;
p2=p2->next;
}
for (int i=0;i<k;i )
{
if (i!=nn)
{
if (sum[i]<min1)
{
min1=sum[i];
distance2=min1;
}
}
}
n1=0;
p2=aa.keyDescriptors2;
while(p2)
{
if (n1==nn)
{
min=100,min1=100;
float *fest_desc2=(float *) malloc(128 * sizeof(float));
fest_desc2=p2->descrip;
// float sum[200];
for (int j=0;j<200;j )
sum[j]=0.0;
p1=aa.keyDescriptors1;
n=0;
while(p1)
{
float *fest_desc1=(float *) malloc(128*sizeof(float));
fest_desc1=p1->descrip;
for (int i=0;i<128;i )
{
dis=fabs(fest_desc2[i]-fest_desc1[i]);
sum[n]=sum[n] dis;
}
if (sum[n]<min)
{
min=sum[n];
distance11=min;
nnn=n;
}
n=n 1;
p1=p1->next;
}
for (int i=0;i<kk;i )
{
if (i!=nnn)
{
if (sum[i]<min1)
{
min1=sum[i];
distance22=min1;
}
}
}
}
n1 ;
p2=p2->next;
}
p1=p3;
//if ((distance1*1.0/distance2<0.4) && (distance11*1.0/distance22<0.4) && fabs(distance1-distance11)<0.000001)
if ((distance1*1.0/distance2<0.4) && (distance11*1.0/distance22<0.4) && (distance1==distance11))
{
n=0;
p2=aa.keyDescriptors2;
while(p2)
{
if (n==nn)
{
xy[m].width1=p1->col;
xy[m].height1=p1->row;
xy[m].width2=p2->col;
xy[m].height2=p2->row;
m ;
count ;
}
n ;
p2=p2->next;
}
}
p1=p1->next;
}
if (m==0)
{
MessageBox("找不到特征点对");
}
else
{
Keyaxis xx;
xx.width1=xy[0].width1;
xx.height1=xy[0].height1;
xx.width2=xy[0].width2;
xx.height2=xy[0].height2;
stitchok(xx);
}
}
void stitchimage::stitch2(IplImage* imag1,IplImage* imag2)
{
p1=aa.keyDescriptors1;
p2=aa.keyDescriptors2;
Keyaxis xy[200];
Keypoint p3=NULL;
float dis;
float distance1,distance11,distance2,distance22;
int k,kk;
k=0;
kk=0;
int x1,y1,x2,y2;
int n,nn,nnn,n1;
int count=0;
int m=0;
float min,min1;
while(p2)
{
k=k 1;
p2=p2->next;
}
while(p1)
{
kk=kk 1;
p1=p1->next;
}
p1=aa.keyDescriptors1;
while(p1)
{
p3=p1;
min=100,min1=100;
float *fest_desc1=(float *) malloc(128 * sizeof(float));
fest_desc1=p1->descrip;
float sum[200];
for (int j=0;j<200;j )
sum[j]=0.0;
n=0;
p2=aa.keyDescriptors2;
while(p2)
{
float *fest_desc2=(float *) malloc(128*sizeof(float));
fest_desc2=p2->descrip;
for (int i=0;i<128;i )
{
dis=fabs(fest_desc1[i]-fest_desc2[i]);
sum[n]=sum[n] dis;
}
if (sum[n]<min)
{
min=sum[n];
distance1=min;
nn=n;
}
n=n 1;
p2=p2->next;
}
for (int i=0;i<k;i )
{
if (i!=nn)
{
if (sum[i]<min1)
{
min1=sum[i];
distance2=min1;
}
}
}
n1=0;
p2=aa.keyDescriptors2;
while(p2)
{
if (n1==nn)
{
min=100,min1=100;
float *fest_desc2=(float *) malloc(128 * sizeof(float));
fest_desc2=p2->descrip;
// float sum[200];
for (int j=0;j<200;j )
sum[j]=0.0;
p1=aa.keyDescriptors1;
n=0;
while(p1)
{
float *fest_desc1=(float *) malloc(128*sizeof(float));
fest_desc1=p1->descrip;
for (int i=0;i<128;i )
{
dis=fabs(fest_desc2[i]-fest_desc1[i]);
sum[n]=sum[n] dis;
}
if (sum[n]<min)
{
min=sum[n];
distance11=min;
nnn=n;
}
n=n 1;
p1=p1->next;
}
for (int i=0;i<kk;i )
{
if (i!=nnn)
{
if (sum[i]<min1)
{
min1=sum[i];
distance22=min1;
}
}
}
}
n1 ;
p2=p2->next;
}
p1=p3;
//if ((distance1*1.0/distance2<0.4) && (distance11*1.0/distance22<0.4) && fabs(distance1-distance11)<0.000001)
if ((distance1*1.0/distance2<0.5) && (distance11*1.0/distance22<0.5) && (distance1==distance11))
{
n=0;
p2=aa.keyDescriptors2;
while(p2)
{
if (n==nn)
{
xy[m].width1=p1->col;
xy[m].height1=p1->row;
xy[m].width2=p2->col;
xy[m].height2=p2->row;
m ;
count ;
}
n ;
p2=p2->next;
}
}
p1=p1->next;
}
if (m==0)
{
MessageBox("找不到特征点对");
}
else
{
Keyaxis xx;
xx.width1=xy[0].width1;
xx.height1=xy[0].height1;
xx.width2=xy[0].width2;
xx.height2=xy[0].height2;
stitchok(xx,imag1,imag2);
}
}
void stitchimage::opencvsolve()
{
int i,j;
float a[3][3],b[3],mm[3][3];
for (i=0;i<3;i )
{
a[i][0]=xy1[i 1].width1,a[i][1]=xy1[i 1].height1,a[i][2]=1;
b[i]=xy1[i 1].width2;
}
//赋值给矩阵
CvMat * A=cvCreateMat(3,3,CV_32FC1);
int nn=A->cols;
float *data=A->data.fl;
for (i=0;i<3;i )
{
for (j=0;j<3;j )
{
data[i*nn j]=a[i][j];
}
}
CvMat * B=cvCreateMat(3,1,CV_32FC1);
int nn1=B->cols;
float *data1=B->data.fl;
for (i=0;i<3;i )
{
data1[i*nn1]=b[i];
}
CvMat * X=cvCreateMat(3,1,CV_32FC1);
cvSolve(A,B,X,CV_LU);
for (i=0;i<3;i )
{
mm[0][i]=cvGet2D(X,i,0).val[0];
}
for (i=0;i<3;i )
{
b[i]=xy1[i 1].height2;
}
//赋值给矩阵
for (i=0;i<3;i )
{
data1[i*nn1]=b[i];
}
cvSolve(A,B,X,CV_LU);
for (i=0;i<3;i )
{
mm[1][i]=cvGet2D(X,i,0).val[0];
}
//求出m6,m7
mm[2][0]=-a[0][1]/(a[0][0]*a[1][1]-a[0][1]*a[1][0]);
mm[2][1]=a[0][0]/(a[0][0]*a[1][1]-a[0][1]*a[1][0]);
mm[2][2]=1;
for (i=0;i<ncount;i )
{
xy1[i].width2=mm[0][0]*(xy1[i].width1) mm[0][1]*(xy1[i].height1) mm[0][2];
xy1[i].height2=mm[1][0]*xy1[i].width1 mm[1][1]*xy1[i].height1 mm[1][2];
}
}
void stitchimage::rotateneighbor(IplImage *image,float angle)
{
int imwidth,imheight;
imwidth=image->width;
imheight=image->height;
//旋转角度，从度转化到弧度
float frotateangle=2*pi*angle/360;
//输入图像四个角坐标，以图像中心为坐标原点
float fsrcx1,fsrcy1,fsrcx2,fsrcy2,fsrcx3,fsrcy3,fsrcx4,fsrcy4;
//旋转后四个角的坐标，以图像中心为坐标原点
float fdstx1,fdsty1,fdstx2,fdsty2,fdstx3,fdsty3,fdstx4,fdsty4;
//计算旋转角度的正弦
float fsina=(float) sin((double) frotateangle);
//计算旋转角度的余弦
float fcosa=(float) cos((double) frotateangle);
//计算原图四个角的坐标
fsrcx1=(float) (-(imwidth-1)/2);
fsrcy1=(float) (imwidth-1)/2;
}
int stitchimage::gauss()
{
int u,p,n=3;
int i,j,k,d=0;
float s=0,t,m,f=0;
float a[3][3],b[3],mm[3][3];
//求出m0，m1，m2
for (i=0;i<3;i )
{
a[i][0]=xy1[i].width1,a[i][1]=xy1[i].height1,a[i][2]=1;
b[i]=xy1[i].width2;
}
for (k=0;k<2;k )
{
for (i=k 1;i<3;i )
{
for (u=k;u<3;u )
{
if (fabs(a[u][k]>s))
{
s=fabs(a[u][k]);
p=u;
}
}
if (p!=k)
{
d=0;
for (j=k;j<3;j )
{
t=a[k][j],a[k][j]=a[p][j],a[p][j]=t;
}
t=b[k];b[k]=b[p];b[p]=t;
}
m=a[i][k]/a[k][k],b[i]=b[i]-m*b[k];
for (j=k 1;j<3;j )
{
a[i][j]=a[i][j]-m*a[k][j];
}
}
}
b[n-1]=b[n-1]/a[n-1][n-1];
for (i=n-2;i>=0;i--)
{
for (j=i 1;j<n;j )
{
f=a[i][j]*b[j];
}
b[i]=(b[i]-f)/a[i][j];
}
for (i=0;i<n;i )
{
mm[0][i]=b[i];
}
//求出m3,m4,m5
n=3;
d=0;
s=0,f=0;
for (i=0;i<3;i )
{
b[i]=xy1[i].height2;
}
for (k=0;k<2;k )
{
for (i=k 1;i<3;i )
{
for (u=k;u<3;u )
{
if (fabs(a[u][k]>s))
{
s=fabs(a[u][k]);
p=u;
}
}
if (p!=k)
{
d=0;
for (j=k;j<3;j )
{
t=a[k][j],a[k][j]=a[p][j],a[p][j]=t;
}
t=b[k];b[k]=b[p];b[p]=t;
}
m=a[i][k]/a[k][k],b[i]=b[i]-m*b[k];
for (j=k 1;j<3;j )
{
a[i][j]=a[i][j]-m*a[k][j];
}
}
}
b[n-1]=b[n-1]/a[n-1][n-1];
for (i=n-2;i>=0;i--)
{
for (j=i 1;j<n;j )
{
f=a[i][j]*b[j];
}
b[i]=(b[i]-f)/a[i][j];
}
for (i=0;i<n;i )
{
mm[1][i]=b[i];
}
//求出m6,m7
mm[2][0]=-a[0][1]/(a[0][0]*a[1][1]-a[0][1]*a[1][0]);
mm[2][1]=a[0][0]/(a[0][0]*a[1][1]-a[0][1]*a[1][0]);
mm[2][2]=1;
//赋值给矩阵
CvMat * M=cvCreateMat(3,3,CV_32FC1);
int nn=M->cols;
float *data=M->data.fl;
for (i=0;i<3;i )
{
for (j=0;j<3;j )
{
data[i*nn j]=mm[i][j];
}
}
for (i=0;i<ncount;i )
{
xy1[i].width2=mm[0][0]*(xy1[i].width1) mm[0][1]*(xy1[i].height1) mm[0][2];
xy1[i].height2=mm[1][0]*xy1[i].width1 mm[1][1]*xy1[i].height1 mm[1][2];
}
return 1;
}
void stitchimage:: stitchok(Keyaxis xx)
{
int width1;
width1=imageptocy11->width;
int x,x1,y,y1,x3,x4,x5,disy;
x=xx.width1;
x1=xx.width2;
x3=width1-x x1;
x4=width1-x3;
x5=width1-x4;
y=xx.height1;
y1=xx.height2;
int dis;
dis=y-y1;
disy=abs(y1-y);
// disy=27;
uchar* data3;
data3=(uchar*)imageptocy11->imageData;
uchar* data4;
data4=(uchar*)imageptocy22->imageData;
int width33,height33;
IplImage *image33=NULL;
CvSize image33_cvsize;
height33=imageptocy11->height-disy;
width33=imageptocy11->width imageptocy22->width-x3;
image33_cvsize.height=height33;
image33_cvsize.width=width33;
uchar *dataout;
image33 = cvCreateImage(image33_cvsize, imageptocy11->depth,imageptocy11->nChannels);
dataout=(uchar*)image33->imageData;
float d,g;
if (dis<0)
{
for(int i=0;i<height33;i )
{
d=1.0;
for(int j=0;j<x4;j )
{
for(int n=0;n<3;n )
dataout[i*image33->widthStep j*image33->nChannels n]=
data3[i*imageptocy11->widthStep j*imageptocy11->nChannels n];
}
for (j=x4;j<width1;j )
{
g=d-1.0/x3*(j-x4);
for(int n=0;n<3;n )
{
dataout[i*image33->widthStep j*image33->nChannels n]=(int)
(g*data3[i*imageptocy11->widthStep j*imageptocy11->nChannels n]
(1.0-g)*data4[(i disy)*imageptocy22->widthStep (j-x4)*imageptocy22->nChannels n]);
}
}
for (j=width1;j<width33;j )
{
for(int n=0;n<3;n )
dataout[i*image33->widthStep j*image33->nChannels n]=
data4[(i disy)*imageptocy22->widthStep (j-x4)*imageptocy22->nChannels n];
}
}
}
else
{
for(int i=disy;i<height33 disy;i )
{
d=1.0;
for(int j=0;j<x4;j )
{
for(int n=0;n<3;n )
dataout[(i-disy)*image33->widthStep j*image33->nChannels n]=
data3[i*imageptocy11->widthStep j*imageptocy11->nChannels n];
}
for (j=x4;j<width1;j )
{
g=d-1.0/x3*(j-x4);
for(int n=0;n<3;n )
{
dataout[(i-disy)*image33->widthStep j*image33->nChannels n]=(int)
(g*data3[i*imageptocy11->widthStep j*imageptocy11->nChannels n]
(1.0-g)*data4[(i-disy)*imageptocy22->widthStep (j-x4)*imageptocy22->nChannels n]);
}
}
for (j=width1;j<width33;j )
{
for(int n=0;n<3;n )
dataout[(i-disy)*image33->widthStep j*image33->nChannels n]=
data4[(i-disy)*imageptocy22->widthStep (j-x4)*imageptocy22->nChannels n];
}
}
}
IplImage* dst1=NULL;
dst1=cvCreateImage(cvGetSize(image33),image33->depth,image33->nChannels);
cvCopy(image33,dst1,NULL);
cvSaveImage("D:/fullimag.jpg",image33);
float wscale = 0.22;
float hscale = 0.19;
CvSize dst1_cvsize;
dst1_cvsize.width = image33->width * wscale; //目标图像的宽为源图象宽的scale倍
dst1_cvsize.height = image33->height * hscale; //目标图像的高为源图象高的scale倍
dst1 = cvCreateImage( dst1_cvsize, image33->depth, image33->nChannels); //构造目标图象
cvResize(image33, dst1, CV_INTER_LINEAR); //缩放源图像到目标图像
cvNamedWindow( "全景图", CV_WINDOW_AUTOSIZE ); //创建用于显示目标图像的窗口
cvShowImage( "全景图", dst1 ); //显示目标图像
cvMoveWindow( "全景图",0,500);
cvWaitKey(-1);
cvReleaseImage(&dst1); //释放目标图像占用的内存
cvDestroyWindow( "全景图" ); //销毁窗口“dst”
}
void stitchimage::Onsiftstitch()
{
// TODO: Add your control notification handler code here
if (DOGoctaves1!=NULL)
{
DOGoctaves1=NULL;
}
if (DOGoctaves2!=NULL)
{
DOGoctaves2=NULL;
}
if (Gaussianpyr1!=NULL)
{
Gaussianpyr1=NULL;
}
if (Gaussianpyr2!=NULL)
{
Gaussianpyr2=NULL;
}
int wd,hd;
wd=imageptocy11->width;
hd=imageptocy11->height;
if (wd >= 1600 && hd >= 1200)
{
stitch1();
}
else
{
stitch2();
}
}
好例子网口号：伸出你的我的手 — 分享！
小贴士
感谢您为本站写下的评论，您的评论对其它用户来说具有重要的参考价值，所以请认真填写。
- 类似“顶”、“沙发”之类没有营养的文字，对勤劳贡献的楼主来说是令人沮丧的反馈信息。
- 相信您也不想看到一排文字/表情墙，所以请不要反馈意义不大的重复字符，也请尽量不要纯表情的回复。
- 提问之前请再仔细看一遍楼主的说明，或许是您遗漏了。
- 请勿到处挖坑绊人、招贴广告。既占空间让人厌烦，又没人会搭理，于人于己都无利。
关于好例子网
本站旨在为广大IT学习爱好者提供一个非营利性互相学习交流分享平台。本站所有资源都可以被免费获取学习研究。本站资源来自网友分享，对搜索内容的合法性不具有预见性、识别性、控制性，仅供学习研究，请务必在下载后24小时内给予删除，不得用于其他任何用途，否则后果自负。基于互联网的特殊性，平台无法对用户传输的作品、信息、内容的权属或合法性、安全性、合规性、真实性、科学性、完整权、有效性等进行实质审查；无论平台是否已进行审查，用户均应自行承担因其传输的作品、信息、内容而可能或已经产生的侵权或权属纠纷等法律责任。本站所有资源不代表本站的观点或立场，基于网友分享，根据中国法律《信息网络传播权保护条例》第二十二与二十三条之规定，若资源存在侵权或相关问题请联系本站客服人员，点此联系我们。关于更多版权及免责申明参见 版权及免责申明
网友评论
我要评论