static void PointsToMat(
const IntrinsicBase * cam,
const PointFeatures & vec_feats,
MatT & m)
{
m.resize(2, vec_feats.size());
typedef typename MatT::Scalar Scalar; // Output matrix type
size_t i = 0;
for( PointFeatures::const_iterator iter = vec_feats.begin();
iter != vec_feats.end(); ++iter, ++i)
{
m.col(i) = cam->get_ud_pixel(Vec2(iter->x(), iter->y()));
}
}
int main() {
const std::string sInputDir = stlplus::folder_up(string(THIS_SOURCE_DIR))
+ "/imageData/SceauxCastle/";
Image<RGBColor> image;
const string jpg_filenameL = sInputDir + "100_7101.jpg";
const string jpg_filenameR = sInputDir + "100_7102.jpg";
Image<unsigned char> imageL, imageR;
ReadImage(jpg_filenameL.c_str(), &imageL);
ReadImage(jpg_filenameR.c_str(), &imageR);
//--
// Detect regions thanks to an image_describer
//--
using namespace openMVG::features;
std::unique_ptr<Image_describer> image_describer(new SIFT_Image_describer);
std::map<IndexT, std::unique_ptr<features::Regions> > regions_perImage;
image_describer->Describe(imageL, regions_perImage[0]);
image_describer->Describe(imageR, regions_perImage[1]);
const SIFT_Regions* regionsL = dynamic_cast<SIFT_Regions*>(regions_perImage.at(0).get());
const SIFT_Regions* regionsR = dynamic_cast<SIFT_Regions*>(regions_perImage.at(1).get());
const PointFeatures
featsL = regions_perImage.at(0)->GetRegionsPositions(),
featsR = regions_perImage.at(1)->GetRegionsPositions();
// Show both images side by side
{
Image<unsigned char> concat;
ConcatH(imageL, imageR, concat);
string out_filename = "01_concat.jpg";
WriteImage(out_filename.c_str(), concat);
}
//- Draw features on the two image (side by side)
{
Image<unsigned char> concat;
ConcatH(imageL, imageR, concat);
//-- Draw features :
for (size_t i=0; i < featsL.size(); ++i ) {
const SIOPointFeature point = regionsL->Features()[i];
DrawCircle(point.x(), point.y(), point.scale(), 255, &concat);
}
for (size_t i=0; i < featsR.size(); ++i ) {
const SIOPointFeature point = regionsR->Features()[i];
DrawCircle(point.x()+imageL.Width(), point.y(), point.scale(), 255, &concat);
}
string out_filename = "02_features.jpg";
WriteImage(out_filename.c_str(), concat);
}
std::vector<IndMatch> vec_PutativeMatches;
//-- Perform matching -> find Nearest neighbor, filtered with Distance ratio
{
// Define a matcher and a metric to find corresponding points
typedef SIFT_Regions::DescriptorT DescriptorT;
typedef L2_Vectorized<DescriptorT::bin_type> Metric;
typedef ArrayMatcherBruteForce<DescriptorT::bin_type, Metric> MatcherT;
// Distance ratio squared due to squared metric
getPutativesMatches<DescriptorT, MatcherT>(
((SIFT_Regions*)regions_perImage.at(0).get())->Descriptors(),
((SIFT_Regions*)regions_perImage.at(1).get())->Descriptors(),
Square(0.8), vec_PutativeMatches);
// Draw correspondences after Nearest Neighbor ratio filter
svgDrawer svgStream( imageL.Width() + imageR.Width(), max(imageL.Height(), imageR.Height()));
svgStream.drawImage(jpg_filenameL, imageL.Width(), imageL.Height());
svgStream.drawImage(jpg_filenameR, imageR.Width(), imageR.Height(), imageL.Width());
for (size_t i = 0; i < vec_PutativeMatches.size(); ++i) {
//Get back linked feature, draw a circle and link them by a line
const SIOPointFeature L = regionsL->Features()[vec_PutativeMatches[i]._i];
const SIOPointFeature R = regionsR->Features()[vec_PutativeMatches[i]._j];
svgStream.drawLine(L.x(), L.y(), R.x()+imageL.Width(), R.y(), svgStyle().stroke("green", 2.0));
svgStream.drawCircle(L.x(), L.y(), L.scale(), svgStyle().stroke("yellow", 2.0));
svgStream.drawCircle(R.x()+imageL.Width(), R.y(), R.scale(),svgStyle().stroke("yellow", 2.0));
}
const string out_filename = "03_siftMatches.svg";
ofstream svgFile( out_filename.c_str() );
svgFile << svgStream.closeSvgFile().str();
svgFile.close();
}
// Fundamental geometry filtering of putative matches
{
//A. get back interest point and send it to the robust estimation framework
Mat xL(2, vec_PutativeMatches.size());
Mat xR(2, vec_PutativeMatches.size());
for (size_t k = 0; k < vec_PutativeMatches.size(); ++k) {
const PointFeature & imaL = featsL[vec_PutativeMatches[k]._i];
const PointFeature & imaR = featsR[vec_PutativeMatches[k]._j];
xL.col(k) = imaL.coords().cast<double>();
xR.col(k) = imaR.coords().cast<double>();
}
//-- Fundamental robust estimation
std::vector<size_t> vec_inliers;
typedef ACKernelAdaptor<
openMVG::fundamental::kernel::SevenPointSolver,
openMVG::fundamental::kernel::SymmetricEpipolarDistanceError,
UnnormalizerT,
Mat3>
KernelType;
KernelType kernel(
xL, imageL.Width(), imageL.Height(),
xR, imageR.Width(), imageR.Height(),
true); // configure as point to line error model.
Mat3 F;
std::pair<double,double> ACRansacOut = ACRANSAC(kernel, vec_inliers, 1024, &F,
4.0, // Upper bound of authorized threshold
true);
const double & thresholdF = ACRansacOut.first;
// Check the fundamental support some point to be considered as valid
if (vec_inliers.size() > KernelType::MINIMUM_SAMPLES *2.5) {
std::cout << "\nFound a fundamental under the confidence threshold of: "
<< thresholdF << " pixels\n\twith: " << vec_inliers.size() << " inliers"
<< " from: " << vec_PutativeMatches.size()
<< " putatives correspondences"
<< std::endl;
//Show fundamental validated point and compute residuals
std::vector<double> vec_residuals(vec_inliers.size(), 0.0);
svgDrawer svgStream( imageL.Width() + imageR.Width(), max(imageL.Height(), imageR.Height()));
svgStream.drawImage(jpg_filenameL, imageL.Width(), imageL.Height());
svgStream.drawImage(jpg_filenameR, imageR.Width(), imageR.Height(), imageL.Width());
for ( size_t i = 0; i < vec_inliers.size(); ++i) {
const SIOPointFeature & LL = regionsL->Features()[vec_PutativeMatches[vec_inliers[i]]._i];
const SIOPointFeature & RR = regionsR->Features()[vec_PutativeMatches[vec_inliers[i]]._j];
const Vec2f L = LL.coords();
const Vec2f R = RR.coords();
svgStream.drawLine(L.x(), L.y(), R.x()+imageL.Width(), R.y(), svgStyle().stroke("green", 2.0));
svgStream.drawCircle(L.x(), L.y(), LL.scale(), svgStyle().stroke("yellow", 2.0));
svgStream.drawCircle(R.x()+imageL.Width(), R.y(), RR.scale(),svgStyle().stroke("yellow", 2.0));
// residual computation
vec_residuals[i] = std::sqrt(KernelType::ErrorT::Error(F,
LL.coords().cast<double>(),
RR.coords().cast<double>()));
}
const string out_filename = "04_ACRansacFundamental.svg";
ofstream svgFile( out_filename.c_str() );
svgFile << svgStream.closeSvgFile().str();
svgFile.close();
// Display some statistics of reprojection errors
float dMin, dMax, dMean, dMedian;
minMaxMeanMedian<float>(vec_residuals.begin(), vec_residuals.end(),
dMin, dMax, dMean, dMedian);
std::cout << std::endl
<< "Fundamental matrix estimation, residuals statistics:" << "\n"
<< "\t-- Residual min:\t" << dMin << std::endl
<< "\t-- Residual median:\t" << dMedian << std::endl
<< "\t-- Residual max:\t " << dMax << std::endl
<< "\t-- Residual mean:\t " << dMean << std::endl;
}
else {
std::cout << "ACRANSAC was unable to estimate a rigid fundamental"
<< std::endl;
}
}
return EXIT_SUCCESS;
}
int main() {
std::cout << "Compute the relative pose between two spherical image."
<< "\nUse an Acontrario robust estimation based on angular errors." << std::endl;
const std::string sInputDir = std::string(THIS_SOURCE_DIR);
const string jpg_filenameL = sInputDir + "/SponzaLion000.jpg";
Image<unsigned char> imageL;
ReadImage(jpg_filenameL.c_str(), &imageL);
const string jpg_filenameR = sInputDir + "/SponzaLion001.jpg";
Image<unsigned char> imageR;
ReadImage(jpg_filenameR.c_str(), &imageR);
//--
// Detect regions thanks to an image_describer
//--
using namespace openMVG::features;
std::unique_ptr<Image_describer> image_describer(new SIFT_Image_describer(SiftParams(-1)));
std::map<IndexT, std::unique_ptr<features::Regions> > regions_perImage;
image_describer->Describe(imageL, regions_perImage[0]);
image_describer->Describe(imageR, regions_perImage[1]);
const SIFT_Regions* regionsL = dynamic_cast<SIFT_Regions*>(regions_perImage.at(0).get());
const SIFT_Regions* regionsR = dynamic_cast<SIFT_Regions*>(regions_perImage.at(1).get());
const PointFeatures
featsL = regions_perImage.at(0)->GetRegionsPositions(),
featsR = regions_perImage.at(1)->GetRegionsPositions();
std::cout << "Left image SIFT count: " << featsL.size() << std::endl;
std::cout << "Right image SIFT count: "<< featsR.size() << std::endl;
// Show both images side by side
{
Image<unsigned char> concat;
ConcatH(imageL, imageR, concat);
string out_filename = "01_concat.jpg";
WriteImage(out_filename.c_str(), concat);
}
//- Draw features on the two image (side by side)
{
Image<unsigned char> concat;
ConcatH(imageL, imageR, concat);
//-- Draw features :
for (size_t i=0; i < featsL.size(); ++i ) {
const SIOPointFeature point = regionsL->Features()[i];
DrawCircle(point.x(), point.y(), point.scale(), 255, &concat);
}
for (size_t i=0; i < featsR.size(); ++i ) {
const SIOPointFeature point = regionsR->Features()[i];
DrawCircle(point.x()+imageL.Width(), point.y(), point.scale(), 255, &concat);
}
string out_filename = "02_features.jpg";
WriteImage(out_filename.c_str(), concat);
}
std::vector<IndMatch> vec_PutativeMatches;
//-- Perform matching -> find Nearest neighbor, filtered with Distance ratio
{
// Find corresponding points
matching::DistanceRatioMatch(
0.8, matching::ANN_L2,
*regions_perImage.at(0).get(),
*regions_perImage.at(1).get(),
vec_PutativeMatches);
IndMatchDecorator<float> matchDeduplicator(vec_PutativeMatches, featsL, featsR);
matchDeduplicator.getDeduplicated(vec_PutativeMatches);
// Draw correspondences after Nearest Neighbor ratio filter
svgDrawer svgStream( imageL.Width() + imageR.Width(), max(imageL.Height(), imageR.Height()));
svgStream.drawImage(jpg_filenameL, imageL.Width(), imageL.Height());
svgStream.drawImage(jpg_filenameR, imageR.Width(), imageR.Height(), imageL.Width());
for (size_t i = 0; i < vec_PutativeMatches.size(); ++i) {
//Get back linked feature, draw a circle and link them by a line
const SIOPointFeature L = regionsL->Features()[vec_PutativeMatches[i].i_];
const SIOPointFeature R = regionsR->Features()[vec_PutativeMatches[i].j_];
svgStream.drawLine(L.x(), L.y(), R.x()+imageL.Width(), R.y(), svgStyle().stroke("green", 2.0));
svgStream.drawCircle(L.x(), L.y(), L.scale(), svgStyle().stroke("yellow", 2.0));
svgStream.drawCircle(R.x()+imageL.Width(), R.y(), R.scale(),svgStyle().stroke("yellow", 2.0));
}
string out_filename = "03_siftMatches.svg";
ofstream svgFile( out_filename.c_str() );
svgFile << svgStream.closeSvgFile().str();
svgFile.close();
}
// Essential geometry filtering of putative matches
{
//A. get back interest point and send it to the robust estimation framework
Mat xL(2, vec_PutativeMatches.size());
Mat xR(2, vec_PutativeMatches.size());
for (size_t k = 0; k < vec_PutativeMatches.size(); ++k) {
const PointFeature & imaL = featsL[vec_PutativeMatches[k].i_];
const PointFeature & imaR = featsR[vec_PutativeMatches[k].j_];
xL.col(k) = imaL.coords().cast<double>();
xR.col(k) = imaR.coords().cast<double>();
}
//-- Convert planar to spherical coordinates
Mat xL_spherical(3,vec_PutativeMatches.size()), xR_spherical(3,vec_PutativeMatches.size());
spherical_cam::planarToSpherical(xL, imageL.Width(), imageL.Height(), xL_spherical);
spherical_cam::planarToSpherical(xR, imageR.Width(), imageR.Height(), xR_spherical);
//-- Essential matrix robust estimation from spherical bearing vectors
{
std::vector<size_t> vec_inliers;
// Use the 8 point solver in order to estimate E
typedef openMVG::spherical_cam::EssentialKernel_spherical Kernel;
// Define the AContrario angular error adaptor
typedef openMVG::robust::ACKernelAdaptor_AngularRadianError<
openMVG::spherical_cam::EightPointRelativePoseSolver,
openMVG::spherical_cam::AngularError,
Mat3>
KernelType;
KernelType kernel(xL_spherical, xR_spherical);
// Robust estimation of the Essential matrix and it's precision
Mat3 E;
const double precision = std::numeric_limits<double>::infinity();
const std::pair<double,double> ACRansacOut =
ACRANSAC(kernel, vec_inliers, 1024, &E, precision, true);
const double & threshold = ACRansacOut.first;
const double & NFA = ACRansacOut.second;
std::cout << "\n Angular threshold found: " << R2D(threshold) << "(Degree)"<<std::endl;
std::cout << "\n #Putatives/#inliers : " << xL_spherical.cols() << "/" << vec_inliers.size() << "\n" << std::endl;
if (vec_inliers.size() > 120)
{
// If an essential matrix have been found
// Extract R|t
// - From the 4 possible solutions extracted from E keep the best
// - (check cheirality of correspondence once triangulation is done)
// Accumulator to find the best solution
std::vector<size_t> f(4, 0);
std::vector<Mat3> Es; // Essential,
std::vector<Mat3> Rs; // Rotation matrix.
std::vector<Vec3> ts; // Translation matrix.
Es.push_back(E);
// Recover best rotation and translation from E.
MotionFromEssential(E, &Rs, &ts);
//-> Test the 4 solutions will all the point
Mat34 P1;
P_From_KRt(Mat3::Identity(), Mat3::Identity(), Vec3::Zero(), &P1);
std::vector< std::vector<size_t> > vec_newInliers(4);
std::vector< std::vector<Vec3> > vec_3D(4);
for (int kk = 0; kk < 4; ++kk) {
const Mat3 &R2 = Rs[kk];
const Vec3 &t2 = ts[kk];
Mat34 P2;
P_From_KRt(Mat3::Identity(), R2, t2, &P2);
//-- For each inlier:
// - triangulate
// - check chierality
for (size_t k = 0; k < vec_inliers.size(); ++k)
{
const Vec3 & x1_ = xL_spherical.col(vec_inliers[k]);
const Vec3 & x2_ = xR_spherical.col(vec_inliers[k]);
//Triangulate
Vec3 X;
openMVG::spherical_cam::TriangulateDLT(P1, x1_, P2, x2_, &X);
//Check positivity of the depth (sign of the dot product)
const Vec3 Mc = R2 * X + t2;
if (x2_.dot(Mc) > 0 && x1_.dot(X) > 0)
{
++f[kk];
vec_newInliers[kk].push_back(vec_inliers[k]);
vec_3D[kk].push_back(X);
}
}
}
std::cout << std::endl << "estimate_Rt_fromE" << std::endl;
std::cout << " #points in front of both cameras for each solution: "
<< f[0] << " " << f[1] << " " << f[2] << " " << f[3] << std::endl;
std::vector<size_t>::iterator iter = max_element(f.begin(), f.end());
if(*iter != 0) {
const size_t index = std::distance(f.begin(),iter);
if (f[index] < 120) {
std::cout << "Found solution have too few 3D points." << std::endl;
}
else {
std::cout << "Export found 3D scene in current directory." << std::endl;
vec_inliers.clear();
vec_inliers = vec_newInliers[index];
std::ostringstream os;
os << "./" << "relativePose_Spherical"<< ".ply";
plyHelper::exportToPly(vec_3D[index], os.str());
}
}
}
}
}
return EXIT_SUCCESS;
}
int main() {
Image<RGBColor> image;
const string jpg_filenameL = stlplus::folder_up(string(THIS_SOURCE_DIR))
+ "/imageData/StanfordMobileVisualSearch/Ace_0.png";
const string jpg_filenameR = stlplus::folder_up(string(THIS_SOURCE_DIR))
+ "/imageData/StanfordMobileVisualSearch/Ace_1.png";
Image<unsigned char> imageL, imageR;
ReadImage(jpg_filenameL.c_str(), &imageL);
ReadImage(jpg_filenameR.c_str(), &imageR);
//--
// Detect regions thanks to an image_describer
//--
using namespace openMVG::features;
std::unique_ptr<Image_describer> image_describer(new SIFT_Image_describer(SiftParams(-1)));
std::map<IndexT, std::unique_ptr<features::Regions> > regions_perImage;
image_describer->Describe(imageL, regions_perImage[0]);
image_describer->Describe(imageR, regions_perImage[1]);
const SIFT_Regions* regionsL = dynamic_cast<SIFT_Regions*>(regions_perImage.at(0).get());
const SIFT_Regions* regionsR = dynamic_cast<SIFT_Regions*>(regions_perImage.at(1).get());
const PointFeatures
featsL = regions_perImage.at(0)->GetRegionsPositions(),
featsR = regions_perImage.at(1)->GetRegionsPositions();
// Show both images side by side
{
Image<unsigned char> concat;
ConcatH(imageL, imageR, concat);
string out_filename = "01_concat.jpg";
WriteImage(out_filename.c_str(), concat);
}
//- Draw features on the two image (side by side)
{
Image<unsigned char> concat;
ConcatH(imageL, imageR, concat);
//-- Draw features :
for (size_t i=0; i < featsL.size(); ++i ) {
const SIOPointFeature point = regionsL->Features()[i];
DrawCircle(point.x(), point.y(), point.scale(), 255, &concat);
}
for (size_t i=0; i < featsR.size(); ++i ) {
const SIOPointFeature point = regionsR->Features()[i];
DrawCircle(point.x()+imageL.Width(), point.y(), point.scale(), 255, &concat);
}
string out_filename = "02_features.jpg";
WriteImage(out_filename.c_str(), concat);
}
std::vector<IndMatch> vec_PutativeMatches;
//-- Perform matching -> find Nearest neighbor, filtered with Distance ratio
{
// Find corresponding points
matching::DistanceRatioMatch(
0.8, matching::BRUTE_FORCE_L2,
*regions_perImage.at(0).get(),
*regions_perImage.at(1).get(),
vec_PutativeMatches);
// Draw correspondences after Nearest Neighbor ratio filter
svgDrawer svgStream( imageL.Width() + imageR.Width(), max(imageL.Height(), imageR.Height()));
svgStream.drawImage(jpg_filenameL, imageL.Width(), imageL.Height());
svgStream.drawImage(jpg_filenameR, imageR.Width(), imageR.Height(), imageL.Width());
for (size_t i = 0; i < vec_PutativeMatches.size(); ++i) {
//Get back linked feature, draw a circle and link them by a line
const SIOPointFeature L = regionsL->Features()[vec_PutativeMatches[i]._i];
const SIOPointFeature R = regionsR->Features()[vec_PutativeMatches[i]._j];
svgStream.drawLine(L.x(), L.y(), R.x()+imageL.Width(), R.y(), svgStyle().stroke("green", 2.0));
svgStream.drawCircle(L.x(), L.y(), L.scale(), svgStyle().stroke("yellow", 2.0));
svgStream.drawCircle(R.x()+imageL.Width(), R.y(), R.scale(),svgStyle().stroke("yellow", 2.0));
}
const std::string out_filename = "03_siftMatches.svg";
std::ofstream svgFile( out_filename.c_str() );
svgFile << svgStream.closeSvgFile().str();
svgFile.close();
}
// Homography geometry filtering of putative matches
{
//A. get back interest point and send it to the robust estimation framework
Mat xL(2, vec_PutativeMatches.size());
Mat xR(2, vec_PutativeMatches.size());
for (size_t k = 0; k < vec_PutativeMatches.size(); ++k) {
const PointFeature & imaL = featsL[vec_PutativeMatches[k]._i];
const PointFeature & imaR = featsR[vec_PutativeMatches[k]._j];
xL.col(k) = imaL.coords().cast<double>();
xR.col(k) = imaR.coords().cast<double>();
}
//-- Homography robust estimation
std::vector<size_t> vec_inliers;
typedef ACKernelAdaptor<
openMVG::homography::kernel::FourPointSolver,
openMVG::homography::kernel::AsymmetricError,
UnnormalizerI,
Mat3>
KernelType;
KernelType kernel(
xL, imageL.Width(), imageL.Height(),
xR, imageR.Width(), imageR.Height(),
false); // configure as point to point error model.
Mat3 H;
std::pair<double,double> ACRansacOut = ACRANSAC(kernel, vec_inliers, 1024, &H,
std::numeric_limits<double>::infinity(),
true);
const double & thresholdH = ACRansacOut.first;
// Check the homography support some point to be considered as valid
if (vec_inliers.size() > KernelType::MINIMUM_SAMPLES *2.5) {
std::cout << "\nFound a homography under the confidence threshold of: "
<< thresholdH << " pixels\n\twith: " << vec_inliers.size() << " inliers"
<< " from: " << vec_PutativeMatches.size()
<< " putatives correspondences"
<< std::endl;
//Show homography validated point and compute residuals
std::vector<double> vec_residuals(vec_inliers.size(), 0.0);
svgDrawer svgStream( imageL.Width() + imageR.Width(), max(imageL.Height(), imageR.Height()));
svgStream.drawImage(jpg_filenameL, imageL.Width(), imageL.Height());
svgStream.drawImage(jpg_filenameR, imageR.Width(), imageR.Height(), imageL.Width());
for ( size_t i = 0; i < vec_inliers.size(); ++i) {
const SIOPointFeature & LL = regionsL->Features()[vec_PutativeMatches[vec_inliers[i]]._i];
const SIOPointFeature & RR = regionsR->Features()[vec_PutativeMatches[vec_inliers[i]]._j];
const Vec2f L = LL.coords();
const Vec2f R = RR.coords();
svgStream.drawLine(L.x(), L.y(), R.x()+imageL.Width(), R.y(), svgStyle().stroke("green", 2.0));
svgStream.drawCircle(L.x(), L.y(), LL.scale(), svgStyle().stroke("yellow", 2.0));
svgStream.drawCircle(R.x()+imageL.Width(), R.y(), RR.scale(),svgStyle().stroke("yellow", 2.0));
// residual computation
vec_residuals[i] = std::sqrt(KernelType::ErrorT::Error(H,
LL.coords().cast<double>(),
RR.coords().cast<double>()));
}
string out_filename = "04_ACRansacHomography.svg";
ofstream svgFile( out_filename.c_str() );
svgFile << svgStream.closeSvgFile().str();
svgFile.close();
// Display some statistics of reprojection errors
float dMin, dMax, dMean, dMedian;
minMaxMeanMedian<float>(vec_residuals.begin(), vec_residuals.end(),
dMin, dMax, dMean, dMedian);
std::cout << std::endl
<< "Homography matrix estimation, residuals statistics:" << "\n"
<< "\t-- Residual min:\t" << dMin << std::endl
<< "\t-- Residual median:\t" << dMedian << std::endl
<< "\t-- Residual max:\t " << dMax << std::endl
<< "\t-- Residual mean:\t " << dMean << std::endl;
//---------------------------------------
// Warp the images to fit the reference view
//---------------------------------------
// reread right image that will be warped to fit left image
ReadImage(jpg_filenameR.c_str(), &image);
WriteImage("query.png", image);
// Create and fill the output image
Image<RGBColor> imaOut(imageL.Width(), imageL.Height());
image::Warp(image, H, imaOut);
const std::string imageNameOut = "query_warped.png";
WriteImage(imageNameOut.c_str(), imaOut);
}
else {
std::cout << "ACRANSAC was unable to estimate a rigid homography"
<< std::endl;
}
}
return EXIT_SUCCESS;
}
int main(int argc, char **argv) {
CmdLine cmd;
std::string sImg1 = stlplus::folder_up(string(THIS_SOURCE_DIR))
+ "/imageData/StanfordMobileVisualSearch/Ace_0.png";
std::string sImg2 = stlplus::folder_up(string(THIS_SOURCE_DIR))
+ "/imageData/StanfordMobileVisualSearch/Ace_1.png";
std::string sOutDir = "./kvldOut";
std::cout << sImg1 << std::endl << sImg2 << std::endl;
cmd.add( make_option('i', sImg1, "img1") );
cmd.add( make_option('j', sImg2, "img2") );
cmd.add( make_option('o', sOutDir, "outdir") );
if (argc > 1)
{
try {
if (argc == 1) throw std::string("Invalid command line parameter.");
cmd.process(argc, argv);
} catch(const std::string& s) {
std::cerr << "Usage: " << argv[0] << ' '
<< "[-i|--img1 file] "
<< "[-j|--img2 file] "
<< "[-o|--outdir path] "
<< std::endl;
std::cerr << s << std::endl;
return EXIT_FAILURE;
}
}
std::cout << " You called : " <<std::endl
<< argv[0] << std::endl
<< "--img1 " << sImg1 << std::endl
<< "--img2 " << sImg2 << std::endl
<< "--outdir " << sOutDir << std::endl;
if (sOutDir.empty()) {
std::cerr << "\nIt is an invalid output directory" << std::endl;
return EXIT_FAILURE;
}
// -----------------------------
// a. List images
// b. Compute features and descriptor
// c. Compute putatives descriptor matches
// d. Geometric filtering of putatives matches
// e. Export some statistics
// -----------------------------
// Create output dir
if (!stlplus::folder_exists(sOutDir))
stlplus::folder_create( sOutDir );
const string jpg_filenameL = sImg1;
const string jpg_filenameR = sImg2;
Image<unsigned char> imageL, imageR;
ReadImage(jpg_filenameL.c_str(), &imageL);
ReadImage(jpg_filenameR.c_str(), &imageR);
//--
// Detect regions thanks to an image_describer
//--
using namespace openMVG::features;
std::unique_ptr<Image_describer> image_describer(new SIFT_Image_describer(SiftParams(-1)));
std::map<IndexT, std::unique_ptr<features::Regions> > regions_perImage;
image_describer->Describe(imageL, regions_perImage[0]);
image_describer->Describe(imageR, regions_perImage[1]);
const SIFT_Regions* regionsL = dynamic_cast<SIFT_Regions*>(regions_perImage.at(0).get());
const SIFT_Regions* regionsR = dynamic_cast<SIFT_Regions*>(regions_perImage.at(1).get());
const PointFeatures
featsL = regions_perImage.at(0)->GetRegionsPositions(),
featsR = regions_perImage.at(1)->GetRegionsPositions();
// Show both images side by side
{
Image<unsigned char> concat;
ConcatH(imageL, imageR, concat);
string out_filename = "00_images.jpg";
WriteImage(out_filename.c_str(), concat);
}
//- Draw features on the two image (side by side)
{
Image<unsigned char> concat;
ConcatH(imageL, imageR, concat);
//-- Draw features :
for (size_t i=0; i < featsL.size(); ++i ) {
const SIOPointFeature point = regionsL->Features()[i];
DrawCircle(point.x(), point.y(), point.scale(), 255, &concat);
}
for (size_t i=0; i < featsR.size(); ++i ) {
const SIOPointFeature point = regionsR->Features()[i];
DrawCircle(point.x()+imageL.Width(), point.y(), point.scale(), 255, &concat);
}
string out_filename = "01_features.jpg";
WriteImage(out_filename.c_str(), concat);
}
std::vector<IndMatch> vec_PutativeMatches;
//-- Perform matching -> find Nearest neighbor, filtered with Distance ratio
{
// Define a matcher and a metric to find corresponding points
typedef SIFT_Regions::DescriptorT DescriptorT;
typedef L2_Vectorized<DescriptorT::bin_type> Metric;
typedef ArrayMatcherBruteForce<DescriptorT::bin_type, Metric> MatcherT;
// Distance ratio squared due to squared metric
getPutativesMatches<DescriptorT, MatcherT>(
((SIFT_Regions*)regions_perImage.at(0).get())->Descriptors(),
((SIFT_Regions*)regions_perImage.at(1).get())->Descriptors(),
Square(0.8), vec_PutativeMatches);
// Draw correspondences after Nearest Neighbor ratio filter
svgDrawer svgStream( imageL.Width() + imageR.Width(), max(imageL.Height(), imageR.Height()));
svgStream.drawImage(jpg_filenameL, imageL.Width(), imageL.Height());
svgStream.drawImage(jpg_filenameR, imageR.Width(), imageR.Height(), imageL.Width());
for (size_t i = 0; i < vec_PutativeMatches.size(); ++i) {
//Get back linked feature, draw a circle and link them by a line
const SIOPointFeature L = regionsL->Features()[vec_PutativeMatches[i]._i];
const SIOPointFeature R = regionsR->Features()[vec_PutativeMatches[i]._j];
svgStream.drawLine(L.x(), L.y(), R.x()+imageL.Width(), R.y(), svgStyle().stroke("green", 2.0));
svgStream.drawCircle(L.x(), L.y(), L.scale(), svgStyle().stroke("yellow", 2.0));
svgStream.drawCircle(R.x()+imageL.Width(), R.y(), R.scale(),svgStyle().stroke("yellow", 2.0));
}
string out_filename = "02_siftMatches.svg";
ofstream svgFile( out_filename.c_str() );
svgFile << svgStream.closeSvgFile().str();
svgFile.close();
}
//K-VLD filter
Image<float> imgA (imageL.GetMat().cast<float>());
Image<float> imgB (imageR.GetMat().cast<float>());
std::vector<Pair> matchesFiltered;
std::vector<Pair> matchesPair;
for (std::vector<IndMatch>::const_iterator i_match = vec_PutativeMatches.begin();
i_match != vec_PutativeMatches.end(); ++i_match){
matchesPair.push_back(std::make_pair(i_match->_i, i_match->_j));
}
std::vector<double> vec_score;
//In order to illustrate the gvld(or vld)-consistant neighbors,
// the following two parameters has been externalized as inputs of the function KVLD.
openMVG::Mat E = openMVG::Mat::Ones(vec_PutativeMatches.size(), vec_PutativeMatches.size())*(-1);
// gvld-consistancy matrix, intitialized to -1, >0 consistancy value, -1=unknow, -2=false
std::vector<bool> valide(vec_PutativeMatches.size(), true);// indices of match in the initial matches, if true at the end of KVLD, a match is kept.
size_t it_num=0;
KvldParameters kvldparameters; // initial parameters of KVLD
while (it_num < 5 &&
kvldparameters.inlierRate > KVLD(imgA, imgB, regionsL->Features(), regionsR->Features(),
matchesPair, matchesFiltered, vec_score,E,valide,kvldparameters)) {
kvldparameters.inlierRate /= 2;
//std::cout<<"low inlier rate, re-select matches with new rate="<<kvldparameters.inlierRate<<std::endl;
kvldparameters.K = 2;
it_num++;
}
std::vector<IndMatch> vec_FilteredMatches;
for (std::vector<Pair>::const_iterator i_matchFilter = matchesFiltered.begin();
i_matchFilter != matchesFiltered.end(); ++i_matchFilter){
vec_FilteredMatches.push_back(IndMatch(i_matchFilter->first, i_matchFilter->second));
}
//Print K-VLD consistent matches
{
svgDrawer svgStream(imageL.Width() + imageR.Width(), max(imageL.Height(), imageR.Height()));
// ".svg"
svgStream.drawImage(jpg_filenameL, imageL.Width(), imageL.Height());
svgStream.drawImage(jpg_filenameR, imageR.Width(), imageR.Height(), imageL.Width());
for (int it1=0; it1<matchesPair.size()-1;it1++){
for (int it2=it1+1; it2<matchesPair.size();it2++){
if (valide[it1] && valide[it2] && E(it1,it2)>=0){
const PointFeature & l1 = featsL[matchesPair[it1].first];
const PointFeature & r1 = featsR[matchesPair[it1].second];
const PointFeature & l2 = featsL[matchesPair[it2].first];
const PointFeature & r2 = featsR[matchesPair[it2].second];
// Compute the width of the current VLD segment
float L = (l1.coords() - l2.coords()).norm();
float width = std::max(1.f, L / (dimension+1.f));
// ".svg"
svgStream.drawLine(l1.x(), l1.y(), l2.x(), l2.y(), svgStyle().stroke("yellow", width));
svgStream.drawLine(r1.x() + imageL.Width(), r1.y(), r2.x() + imageL.Width(), r2.y(), svgStyle().stroke("yellow", width));
}
}
}
string out_filename = "05_KVLD_Matches.svg";
out_filename = stlplus::create_filespec(sOutDir, out_filename);
ofstream svgFile( out_filename.c_str() );
svgFile << svgStream.closeSvgFile().str();
svgFile.close();
}
{
//Print keypoints kept by K-VLD
svgDrawer svgStream(imageL.Width() + imageR.Width(), max(imageL.Height(), imageR.Height()));
// ".svg"
svgStream.drawImage(jpg_filenameL, imageL.Width(), imageL.Height());
svgStream.drawImage(jpg_filenameR, imageR.Width(), imageR.Height(), imageL.Width());
for (int it=0; it<matchesPair.size();it++){
if (valide[it]){
const PointFeature & l = featsL[matchesPair[it].first];
const PointFeature & r = featsR[matchesPair[it].second];
// ".svg"
svgStream.drawCircle(l.x(), l.y(), 10, svgStyle().stroke("yellow", 2.0));
svgStream.drawCircle(r.x() + imageL.Width(), r.y(), 10, svgStyle().stroke("yellow", 2.0));
}
}
string out_filename = "06_KVLD_Keypoints.svg";
out_filename = stlplus::create_filespec(sOutDir, out_filename);
ofstream svgFile( out_filename.c_str() );
svgFile << svgStream.closeSvgFile().str();
svgFile.close();
}
Image <unsigned char> imageOutL = imageL;
Image <unsigned char> imageOutR = imageR;
getKVLDMask(
&imageOutL, &imageOutR,
regionsL->Features(), regionsR->Features(),
matchesPair,
valide,
E);
{
string out_filename = "07_Left-K-VLD-MASK.jpg";
out_filename = stlplus::create_filespec(sOutDir, out_filename);
WriteImage(out_filename.c_str(), imageOutL);
}
{
string out_filename = "08_Right-K-VLD-MASK.jpg";
out_filename = stlplus::create_filespec(sOutDir, out_filename);
WriteImage(out_filename.c_str(), imageOutR);
}
return EXIT_SUCCESS;
}
int main() {
const std::string sInputDir = stlplus::folder_up(string(THIS_SOURCE_DIR))
+ "/imageData/SceauxCastle/";
const string jpg_filenameL = sInputDir + "100_7101.jpg";
const string jpg_filenameR = sInputDir + "100_7102.jpg";
Image<unsigned char> imageL, imageR;
ReadImage(jpg_filenameL.c_str(), &imageL);
ReadImage(jpg_filenameR.c_str(), &imageR);
//--
// Detect regions thanks to an image_describer
//--
using namespace openMVG::features;
std::unique_ptr<Image_describer> image_describer(new SIFT_Image_describer);
std::map<IndexT, std::unique_ptr<features::Regions> > regions_perImage;
image_describer->Describe(imageL, regions_perImage[0]);
image_describer->Describe(imageR, regions_perImage[1]);
const SIFT_Regions* regionsL = dynamic_cast<SIFT_Regions*>(regions_perImage.at(0).get());
const SIFT_Regions* regionsR = dynamic_cast<SIFT_Regions*>(regions_perImage.at(1).get());
const PointFeatures
featsL = regions_perImage.at(0)->GetRegionsPositions(),
featsR = regions_perImage.at(1)->GetRegionsPositions();
// Show both images side by side
{
Image<unsigned char> concat;
ConcatH(imageL, imageR, concat);
string out_filename = "01_concat.jpg";
WriteImage(out_filename.c_str(), concat);
}
//- Draw features on the two image (side by side)
{
Image<unsigned char> concat;
ConcatH(imageL, imageR, concat);
//-- Draw features :
for (size_t i=0; i < featsL.size(); ++i ) {
const SIOPointFeature point = regionsL->Features()[i];
DrawCircle(point.x(), point.y(), point.scale(), 255, &concat);
}
for (size_t i=0; i < featsR.size(); ++i ) {
const SIOPointFeature point = regionsR->Features()[i];
DrawCircle(point.x()+imageL.Width(), point.y(), point.scale(), 255, &concat);
}
string out_filename = "02_features.jpg";
WriteImage(out_filename.c_str(), concat);
}
std::vector<IndMatch> vec_PutativeMatches;
//-- Perform matching -> find Nearest neighbor, filtered with Distance ratio
{
// Define a matcher and a metric to find corresponding points
typedef SIFT_Regions::DescriptorT DescriptorT;
typedef L2_Vectorized<DescriptorT::bin_type> Metric;
typedef ArrayMatcherBruteForce<DescriptorT::bin_type, Metric> MatcherT;
// Distance ratio squared due to squared metric
getPutativesMatches<DescriptorT, MatcherT>(
((SIFT_Regions*)regions_perImage.at(0).get())->Descriptors(),
((SIFT_Regions*)regions_perImage.at(1).get())->Descriptors(),
Square(0.8), vec_PutativeMatches);
IndMatchDecorator<float> matchDeduplicator(
vec_PutativeMatches, featsL, featsR);
matchDeduplicator.getDeduplicated(vec_PutativeMatches);
std::cout
<< regions_perImage.at(0)->RegionCount() << " #Features on image A" << std::endl
<< regions_perImage.at(1)->RegionCount() << " #Features on image B" << std::endl
<< vec_PutativeMatches.size() << " #matches with Distance Ratio filter" << std::endl;
// Draw correspondences after Nearest Neighbor ratio filter
svgDrawer svgStream( imageL.Width() + imageR.Width(), max(imageL.Height(), imageR.Height()));
svgStream.drawImage(jpg_filenameL, imageL.Width(), imageL.Height());
svgStream.drawImage(jpg_filenameR, imageR.Width(), imageR.Height(), imageL.Width());
for (size_t i = 0; i < vec_PutativeMatches.size(); ++i) {
//Get back linked feature, draw a circle and link them by a line
const SIOPointFeature L = regionsL->Features()[vec_PutativeMatches[i]._i];
const SIOPointFeature R = regionsR->Features()[vec_PutativeMatches[i]._j];
svgStream.drawLine(L.x(), L.y(), R.x()+imageL.Width(), R.y(), svgStyle().stroke("green", 2.0));
svgStream.drawCircle(L.x(), L.y(), L.scale(), svgStyle().stroke("yellow", 2.0));
svgStream.drawCircle(R.x()+imageL.Width(), R.y(), R.scale(),svgStyle().stroke("yellow", 2.0));
}
const std::string out_filename = "03_siftMatches.svg";
std::ofstream svgFile( out_filename.c_str() );
svgFile << svgStream.closeSvgFile().str();
svgFile.close();
}
// Essential geometry filtering of putative matches
{
Mat3 K;
//read K from file
if (!readIntrinsic(stlplus::create_filespec(sInputDir,"K","txt"), K))
{
std::cerr << "Cannot read intrinsic parameters." << std::endl;
return EXIT_FAILURE;
}
//A. prepare the corresponding putatives points
Mat xL(2, vec_PutativeMatches.size());
Mat xR(2, vec_PutativeMatches.size());
for (size_t k = 0; k < vec_PutativeMatches.size(); ++k) {
const PointFeature & imaL = featsL[vec_PutativeMatches[k]._i];
const PointFeature & imaR = featsR[vec_PutativeMatches[k]._j];
xL.col(k) = imaL.coords().cast<double>();
xR.col(k) = imaR.coords().cast<double>();
}
//B. Compute the relative pose thanks to a essential matrix estimation
std::pair<size_t, size_t> size_imaL(imageL.Width(), imageL.Height());
std::pair<size_t, size_t> size_imaR(imageR.Width(), imageR.Height());
sfm::RelativePose_Info relativePose_info;
if (!sfm::robustRelativePose(K, K, xL, xR, relativePose_info, size_imaL, size_imaR, 256))
{
std::cerr << " /!\\ Robust relative pose estimation failure."
<< std::endl;
return EXIT_FAILURE;
}
std::cout << "\nFound an Essential matrix:\n"
<< "\tprecision: " << relativePose_info.found_residual_precision << " pixels\n"
<< "\t#inliers: " << relativePose_info.vec_inliers.size() << "\n"
<< "\t#matches: " << vec_PutativeMatches.size()
<< std::endl;
// Show Essential validated point
svgDrawer svgStream( imageL.Width() + imageR.Width(), max(imageL.Height(), imageR.Height()));
svgStream.drawImage(jpg_filenameL, imageL.Width(), imageL.Height());
svgStream.drawImage(jpg_filenameR, imageR.Width(), imageR.Height(), imageL.Width());
for (size_t i = 0; i < relativePose_info.vec_inliers.size(); ++i) {
const SIOPointFeature & LL = regionsL->Features()[vec_PutativeMatches[relativePose_info.vec_inliers[i]]._i];
const SIOPointFeature & RR = regionsR->Features()[vec_PutativeMatches[relativePose_info.vec_inliers[i]]._j];
const Vec2f L = LL.coords();
const Vec2f R = RR.coords();
svgStream.drawLine(L.x(), L.y(), R.x()+imageL.Width(), R.y(), svgStyle().stroke("green", 2.0));
svgStream.drawCircle(L.x(), L.y(), LL.scale(), svgStyle().stroke("yellow", 2.0));
svgStream.drawCircle(R.x()+imageL.Width(), R.y(), RR.scale(),svgStyle().stroke("yellow", 2.0));
}
const std::string out_filename = "04_ACRansacEssential.svg";
std::ofstream svgFile( out_filename.c_str() );
svgFile << svgStream.closeSvgFile().str();
svgFile.close();
//C. Triangulate and export inliers as a PLY scene
std::vector<Vec3> vec_3DPoints;
// Setup camera intrinsic and poses
Pinhole_Intrinsic intrinsic0(imageL.Width(), imageL.Height(), K(0, 0), K(0, 2), K(1, 2));
Pinhole_Intrinsic intrinsic1(imageR.Width(), imageR.Height(), K(0, 0), K(0, 2), K(1, 2));
const Pose3 pose0 = Pose3(Mat3::Identity(), Vec3::Zero());
const Pose3 pose1 = relativePose_info.relativePose;
// Init structure by inlier triangulation
const Mat34 P1 = intrinsic0.get_projective_equivalent(pose0);
const Mat34 P2 = intrinsic1.get_projective_equivalent(pose1);
std::vector<double> vec_residuals;
vec_residuals.reserve(relativePose_info.vec_inliers.size() * 4);
for (size_t i = 0; i < relativePose_info.vec_inliers.size(); ++i) {
const SIOPointFeature & LL = regionsL->Features()[vec_PutativeMatches[relativePose_info.vec_inliers[i]]._i];
const SIOPointFeature & RR = regionsR->Features()[vec_PutativeMatches[relativePose_info.vec_inliers[i]]._j];
// Point triangulation
Vec3 X;
TriangulateDLT(P1, LL.coords().cast<double>(), P2, RR.coords().cast<double>(), &X);
// Reject point that is behind the camera
if (pose0.depth(X) < 0 && pose1.depth(X) < 0)
continue;
const Vec2 residual0 = intrinsic0.residual(pose0, X, LL.coords().cast<double>());
const Vec2 residual1 = intrinsic1.residual(pose1, X, RR.coords().cast<double>());
vec_residuals.push_back(fabs(residual0(0)));
vec_residuals.push_back(fabs(residual0(1)));
vec_residuals.push_back(fabs(residual1(0)));
vec_residuals.push_back(fabs(residual1(1)));
}
// Display some statistics of reprojection errors
float dMin, dMax, dMean, dMedian;
minMaxMeanMedian<float>(vec_residuals.begin(), vec_residuals.end(),
dMin, dMax, dMean, dMedian);
std::cout << std::endl
<< "Triangulation residuals statistics:" << "\n"
<< "\t-- Residual min:\t" << dMin << "\n"
<< "\t-- Residual median:\t" << dMedian << "\n"
<< "\t-- Residual max:\t " << dMax << "\n"
<< "\t-- Residual mean:\t " << dMean << std::endl;
// Export as PLY (camera pos + 3Dpoints)
std::vector<Vec3> vec_camPos;
vec_camPos.push_back( pose0.center() );
vec_camPos.push_back( pose1.center() );
exportToPly(vec_3DPoints, vec_camPos, "EssentialGeometry.ply");
}
return EXIT_SUCCESS;
}
int main() {
Image<RGBColor> image;
const string jpg_filenameL = stlplus::folder_up(string(THIS_SOURCE_DIR))
+ "/imageData/StanfordMobileVisualSearch/Ace_0.png";
const string jpg_filenameR = stlplus::folder_up(string(THIS_SOURCE_DIR))
+ "/imageData/StanfordMobileVisualSearch/Ace_1.png";
Image<unsigned char> imageL, imageR;
ReadImage(jpg_filenameL.c_str(), &imageL);
ReadImage(jpg_filenameR.c_str(), &imageR);
//--
// Detect regions thanks to an image_describer
//--
using namespace openMVG::features;
std::unique_ptr<Image_describer> image_describer(new SIFT_Image_describer(SiftParams(-1)));
std::map<IndexT, std::unique_ptr<features::Regions> > regions_perImage;
image_describer->Describe(imageL, regions_perImage[0]);
image_describer->Describe(imageR, regions_perImage[1]);
const SIFT_Regions* regionsL = dynamic_cast<SIFT_Regions*>(regions_perImage.at(0).get());
const SIFT_Regions* regionsR = dynamic_cast<SIFT_Regions*>(regions_perImage.at(1).get());
const PointFeatures
featsL = regions_perImage.at(0)->GetRegionsPositions(),
featsR = regions_perImage.at(1)->GetRegionsPositions();
// Show both images side by side
{
Image<unsigned char> concat;
ConcatH(imageL, imageR, concat);
string out_filename = "01_concat.jpg";
WriteImage(out_filename.c_str(), concat);
}
//- Draw features on the two image (side by side)
{
Image<unsigned char> concat;
ConcatH(imageL, imageR, concat);
//-- Draw features :
for (size_t i=0; i < featsL.size(); ++i ) {
const SIOPointFeature point = regionsL->Features()[i];
DrawCircle(point.x(), point.y(), point.scale(), 255, &concat);
}
for (size_t i=0; i < featsR.size(); ++i ) {
const SIOPointFeature point = regionsR->Features()[i];
DrawCircle(point.x()+imageL.Width(), point.y(), point.scale(), 255, &concat);
}
string out_filename = "02_features.jpg";
WriteImage(out_filename.c_str(), concat);
}
std::vector<IndMatch> vec_PutativeMatches;
//-- Perform matching -> find Nearest neighbor, filtered with Distance ratio
{
// Define a matcher and a metric to find corresponding points
typedef SIFT_Regions::DescriptorT DescriptorT;
typedef L2_Vectorized<DescriptorT::bin_type> Metric;
typedef ArrayMatcherBruteForce<DescriptorT::bin_type, Metric> MatcherT;
// Distance ratio squared due to squared metric
getPutativesMatches<DescriptorT, MatcherT>(
((SIFT_Regions*)regions_perImage.at(0).get())->Descriptors(),
((SIFT_Regions*)regions_perImage.at(1).get())->Descriptors(),
Square(0.8), vec_PutativeMatches);
// Draw correspondences after Nearest Neighbor ratio filter
svgDrawer svgStream( imageL.Width() + imageR.Width(), max(imageL.Height(), imageR.Height()));
svgStream.drawImage(jpg_filenameL, imageL.Width(), imageL.Height());
svgStream.drawImage(jpg_filenameR, imageR.Width(), imageR.Height(), imageL.Width());
for (size_t i = 0; i < vec_PutativeMatches.size(); ++i) {
//Get back linked feature, draw a circle and link them by a line
const SIOPointFeature L = regionsL->Features()[vec_PutativeMatches[i]._i];
const SIOPointFeature R = regionsR->Features()[vec_PutativeMatches[i]._j];
svgStream.drawLine(L.x(), L.y(), R.x()+imageL.Width(), R.y(), svgStyle().stroke("green", 2.0));
svgStream.drawCircle(L.x(), L.y(), L.scale(), svgStyle().stroke("yellow", 2.0));
svgStream.drawCircle(R.x()+imageL.Width(), R.y(), R.scale(),svgStyle().stroke("yellow", 2.0));
}
string out_filename = "03_siftMatches.svg";
ofstream svgFile( out_filename.c_str() );
svgFile << svgStream.closeSvgFile().str();
svgFile.close();
}
// Homography geometry filtering of putative matches
{
//A. get back interest point and send it to the robust estimation framework
Mat xL(2, vec_PutativeMatches.size());
Mat xR(2, vec_PutativeMatches.size());
for (size_t k = 0; k < vec_PutativeMatches.size(); ++k) {
const PointFeature & imaL = featsL[vec_PutativeMatches[k]._i];
const PointFeature & imaR = featsR[vec_PutativeMatches[k]._j];
xL.col(k) = imaL.coords().cast<double>();
xR.col(k) = imaR.coords().cast<double>();
}
//-- Homography robust estimation
std::vector<size_t> vec_inliers;
typedef ACKernelAdaptor<
openMVG::homography::kernel::FourPointSolver,
openMVG::homography::kernel::AsymmetricError,
UnnormalizerI,
Mat3>
KernelType;
KernelType kernel(
xL, imageL.Width(), imageL.Height(),
xR, imageR.Width(), imageR.Height(),
false); // configure as point to point error model.
Mat3 H;
std::pair<double,double> ACRansacOut = ACRANSAC(kernel, vec_inliers, 1024, &H,
std::numeric_limits<double>::infinity(),
true);
const double & thresholdH = ACRansacOut.first;
// Check the homography support some point to be considered as valid
if (vec_inliers.size() > KernelType::MINIMUM_SAMPLES *2.5) {
std::cout << "\nFound a homography under the confidence threshold of: "
<< thresholdH << " pixels\n\twith: " << vec_inliers.size() << " inliers"
<< " from: " << vec_PutativeMatches.size()
<< " putatives correspondences"
<< std::endl;
//Show homography validated point and compute residuals
std::vector<double> vec_residuals(vec_inliers.size(), 0.0);
svgDrawer svgStream( imageL.Width() + imageR.Width(), max(imageL.Height(), imageR.Height()));
svgStream.drawImage(jpg_filenameL, imageL.Width(), imageL.Height());
svgStream.drawImage(jpg_filenameR, imageR.Width(), imageR.Height(), imageL.Width());
for ( size_t i = 0; i < vec_inliers.size(); ++i) {
const SIOPointFeature & LL = regionsL->Features()[vec_PutativeMatches[vec_inliers[i]]._i];
const SIOPointFeature & RR = regionsR->Features()[vec_PutativeMatches[vec_inliers[i]]._j];
const Vec2f L = LL.coords();
const Vec2f R = RR.coords();
svgStream.drawLine(L.x(), L.y(), R.x()+imageL.Width(), R.y(), svgStyle().stroke("green", 2.0));
svgStream.drawCircle(L.x(), L.y(), LL.scale(), svgStyle().stroke("yellow", 2.0));
svgStream.drawCircle(R.x()+imageL.Width(), R.y(), RR.scale(),svgStyle().stroke("yellow", 2.0));
// residual computation
vec_residuals[i] = std::sqrt(KernelType::ErrorT::Error(H,
LL.coords().cast<double>(),
RR.coords().cast<double>()));
}
string out_filename = "04_ACRansacHomography.svg";
ofstream svgFile( out_filename.c_str() );
svgFile << svgStream.closeSvgFile().str();
svgFile.close();
// Display some statistics of reprojection errors
float dMin, dMax, dMean, dMedian;
minMaxMeanMedian<float>(vec_residuals.begin(), vec_residuals.end(),
dMin, dMax, dMean, dMedian);
std::cout << std::endl
<< "Homography matrix estimation, residuals statistics:" << "\n"
<< "\t-- Residual min:\t" << dMin << std::endl
<< "\t-- Residual median:\t" << dMedian << std::endl
<< "\t-- Residual max:\t " << dMax << std::endl
<< "\t-- Residual mean:\t " << dMean << std::endl;
// --
// Perform GUIDED MATCHING
// --
// Use the computed model to check valid correspondences
// a. by considering only the geometric error,
// b. by considering geometric error and descriptor distance ratio.
std::vector< std::vector<IndMatch> > vec_corresponding_indexes(2);
Mat xL, xR;
PointsToMat(featsL, xL);
PointsToMat(featsR, xR);
//a. by considering only the geometric error
geometry_aware::GuidedMatching<Mat3, openMVG::homography::kernel::AsymmetricError>(
H, xL, xR, Square(thresholdH), vec_corresponding_indexes[0]);
std::cout << "\nGuided homography matching (geometric error) found "
<< vec_corresponding_indexes[0].size() << " correspondences."
<< std::endl;
// b. by considering geometric error and descriptor distance ratio
typedef SIFT_Regions::DescriptorT DescriptorT;
geometry_aware::GuidedMatching
<Mat3, openMVG::homography::kernel::AsymmetricError,
DescriptorT, L2_Vectorized<DescriptorT::bin_type> >(
H,
xL, ((SIFT_Regions*)regions_perImage.at(0).get())->Descriptors(),
xR, ((SIFT_Regions*)regions_perImage.at(1).get())->Descriptors(),
Square(thresholdH), Square(0.8),
vec_corresponding_indexes[1]);
std::cout << "\nGuided homography matching "
<< "(geometric + descriptor distance ratio) found "
<< vec_corresponding_indexes[1].size() << " correspondences."
<< std::endl;
for (size_t idx = 0; idx < 2; ++idx)
{
const std::vector<IndMatch> & vec_corresponding_index = vec_corresponding_indexes[idx];
//Show homography validated correspondences
svgDrawer svgStream( imageL.Width() + imageR.Width(), max(imageL.Height(), imageR.Height()));
svgStream.drawImage(jpg_filenameL, imageL.Width(), imageL.Height());
svgStream.drawImage(jpg_filenameR, imageR.Width(), imageR.Height(), imageL.Width());
for ( size_t i = 0; i < vec_corresponding_index.size(); ++i) {
const SIOPointFeature & LL = regionsL->Features()[vec_corresponding_index[i]._i];
const SIOPointFeature & RR = regionsR->Features()[vec_corresponding_index[i]._j];
const Vec2f L = LL.coords();
const Vec2f R = RR.coords();
svgStream.drawLine(L.x(), L.y(), R.x()+imageL.Width(), R.y(), svgStyle().stroke("green", 2.0));
svgStream.drawCircle(L.x(), L.y(), LL.scale(), svgStyle().stroke("yellow", 2.0));
svgStream.drawCircle(R.x()+imageL.Width(), R.y(), RR.scale(),svgStyle().stroke("yellow", 2.0));
}
const string out_filename =
(idx == 0) ? "04_ACRansacHomography_guided_geom.svg"
: "04_ACRansacHomography_guided_geom_distratio.svg";
ofstream svgFile( out_filename.c_str() );
svgFile << svgStream.closeSvgFile().str();
svgFile.close();
}
}
else {
std::cout << "ACRANSAC was unable to estimate a rigid homography"
<< std::endl;
}
}
return EXIT_SUCCESS;
}
