static void create_bone_model_(float radius, float ratio, float length, vtx::fvtxs& vs, vtx::fvtxs& ns)
{
static vtx::fvtx rig[4];
float r = radius * 0.707f;
float w = length * ratio;
rig[0].set( r, r, w);
rig[1].set( r, -r, w);
rig[2].set(-r, -r, w);
rig[3].set(-r, r, w);
vtx::fvtx end(0.0f, 0.0f, length);
for(int i = 0; i < 4; ++i) {
vs.push_back(vtx::fvtx(0.0f));
vs.push_back(rig[i]);
vs.push_back(rig[(i + 1) & 3]);
{
vtx::fvtx n;
outer_product(rig[i], rig[(i + 1) & 3], n);
ns.push_back(n);
ns.push_back(n);
ns.push_back(n);
}
vs.push_back(rig[(i + 1) & 3]);
vs.push_back(rig[i]);
vs.push_back(end);
{
vtx::fvtx n;
outer_product(rig[i + 1] - end, rig[i] - end, n);
ns.push_back(n);
ns.push_back(n);
ns.push_back(n);
}
}
}
void Puckering::CalculateGradZ(const vector<double> &X)
{
/* Calculate puckering variable z_j */
this->CalculateZ(X);
int iN, jN, iD;
fill(_dR1.begin(), _dR1.end(), 0.0);
fill(_dR2.begin(), _dR2.end(), 0.0);
/* Calculate derivatives of R' and R'' */
for ( iN = 0; iN < _nAtoms; ++iN )
for ( jN = 0; jN < _nAtoms; ++jN )
{
_dR1.at(iN) += _Delta.at(iN*_nAtoms+jN)*_sinx.at(jN);
_dR2.at(iN) += _Delta.at(iN*_nAtoms+jN)*_cosx.at(jN);
}
/* Calculate some dot products */
_R1R1 = inner_product(_R1.begin(), _R1.end(), _R1.begin(), 0.0);
_R2R2 = inner_product(_R2.begin(), _R2.end(), _R2.begin(), 0.0);
_R1R2 = inner_product(_R1.begin(), _R1.end(), _R2.begin(), 0.0);
/* Calculate some cross products */
_R1R1 = inner_product(_R1.begin(), _R1.end(), _R1.begin(), 0.0);
int jj = 0;
for ( jN = 0; jN < _nAtoms; ++jN )
{
jj = jN*_sdim;
outer_product(_R2.begin() , _R .begin()+jj, _R2xRj.begin()+jj);
outer_product(_R .begin()+jj, _R1.begin() , _RjxR1.begin()+jj);
}
/* Calculate derivativs of z_j */
double norm2Inv = 0.5/(_norm*_norm);
double normInv = 1.0/_norm;
for ( iD = 0; iD < _sdim; ++iD )
for ( iN = 0; iN < _nAtoms; ++iN )
{
_temp1 = 2.0 * ( _dR1.at(iN)*( _R1.at(iD)*_R2R2 - _R2.at(iD)*_R1R2 )
+_dR2.at(iN)*( _R2.at(iD)*_R1R1 - _R1.at(iD)*_R1R2 ) );
for ( jN = 0; jN < _nAtoms; ++jN )
{
_temp2 = _Delta.at(iN*_nAtoms+jN) *_R1xR2.at(iD)
+ _dR1.at(iN)*_R2xRj.at(jN*_sdim+iD)
+ _dR2.at(iN)*_RjxR1.at(jN*_sdim+iD);
// effective computation of grad_i z_j
_dz.at(iN*_hdim+jN*_sdim+iD) = _temp2*normInv -_z.at(jN)*_temp1*norm2Inv ;
}
}
}
// sample user nodes hyperprior
// according to equation A.4 in Xiong paper
void sample_V(){
vec Vmean = zeros(D);
mat VVT = calc_MMT(M, M+N, Vmean);
double beta0_ = beta0[0] + N;
vec mu0_ = (N*Vmean)/beta0_;
double nu0_ = nu0 +N;
vec dMu = - Vmean;
if (debug)
std::cout<<"dMu:"<<dMu<<"beta0: "<<beta0[0]<<" beta0_ "<<beta0_<<" nu0_ " <<nu0_<<endl;
mat VmeanT = N*outer_product(Vmean, Vmean);
assert(VmeanT.rows() == D && VmeanT.cols() == D);
mat dMuT = (beta0[0]/beta0_)*VmeanT;
mat iW0_ = iW0 + VVT - VmeanT + dMuT;
mat W0_;
bool ret = inv(iW0_, W0_);
assert(ret);
mat tmp = (W0_+transpose(W0_))*0.5;
if (debug)
std::cout<<"iW0: "<<iW0<<" VVT: "<<VVT<<" VmeanT: "<<VmeanT<<" dMuT: " <<dMuT<<"W0_"<< W0_<<" tmp: " << tmp<<" nu0_: "<<nu0_<<endl;
A_V = wishrnd(tmp, nu0_);
mat tmp2;
ret = inv(beta0_*A_V, tmp2);
assert(ret);
mu_V = mvnrndex(mu0_, tmp2, D, 0);
if (debug)
std::cout<<"Sampling from V: A_V" <<A_V<<" mu_V: "<<mu_V<<" Vmean: "<<Vmean<<" W0_: "<<W0_<<" tmp: "<<tmp<<endl;
}
// sample movie nodes hyperprior
// according to equation A.3 in Xiong paper.
void sample_U(){
vec Umean = zeros(D);
mat UUT = calc_MMT(0,M,Umean);
double beta0_ = beta0[0] + M;
vec mu0_ = (M*Umean)/beta0_;
double nu0_ = nu0 +M;
vec dMu = - Umean;
if (debug)
std::cout<<"dMu:"<<dMu<<"beta0: "<<beta0[0]<<" beta0_ "<<beta0_<<" nu0_ " <<nu0_<<" mu0_ " << mu0_<<endl;
mat UmeanT = M*outer_product(Umean, Umean);
assert(UmeanT.rows() == D && UmeanT.cols() == D);
mat dMuT = (beta0[0]/beta0_)*UmeanT;
mat iW0_ = iW0 + UUT - UmeanT + dMuT;
mat W0_;
bool ret =inv(iW0_, W0_);
assert(ret);
mat tmp = (W0_+transpose(W0_))*0.5;
if (debug)
std::cout<<iW0<<UUT<<UmeanT<<dMuT<<W0_<<tmp<<nu0_<<endl;
A_U = wishrnd(tmp, nu0_);
mat tmp2;
ret = inv(beta0_ * A_U, tmp2);
assert(ret);
mu_U = mvnrndex(mu0_, tmp2, D, 0);
if (debug)
std::cout<<"Sampling from U" <<A_U<<" "<<mu_U<<" "<<Umean<<" "<<W0_<<tmp<<endl;
}
void Puckering::CalculateZ(const vector<double> &X)
{
int jN, iN, iD;
/* Coordinates of ring elements with respect to the geometrical center */
fill(_R.begin(), _R.end(), 0.0);
for ( jN = 0; jN < _nAtoms; ++jN)
for (iD = 0; iD < _sdim; ++iD)
for ( iN = 0; iN < _nAtoms; ++iN)
_R.at(jN*_sdim+iD) += X.at(iN*_sdim+iD)*_Delta.at(iN*_nAtoms+jN);
fill(_R1.begin(), _R1.end(), 0.0);
fill(_R2.begin(), _R2.end(), 0.0);
/* R1 = sinx*R and R2 = sinx*R */
for (iD = 0; iD < _sdim; ++iD)
for ( jN = 0; jN < _nAtoms; ++jN)
{
_R1.at(iD) += _sinx.at(jN)*_R.at(jN*_sdim+iD);
_R2.at(iD) += _cosx.at(jN)*_R.at(jN*_sdim+iD);
}
/* R1xR2 : Cross product of R1 and R2 */
outer_product(_R1.begin(), _R2.begin(), _R1xR2.begin());
/* |R1xR2| : L2 norm of cross product of R1 and R2 */
_norm = sqrt( _R1xR2.at(0)*_R1xR2.at(0) + _R1xR2.at(1)*_R1xR2.at(1) + _R1xR2.at(2)*_R1xR2.at(2));
/* Normal vector to the mean plane is n = R1xR2/|R1xR2|,
* so z = R.n */
for ( jN = 0; jN < _nAtoms; ++jN)
_z.at(jN) = inner_product( _R1xR2.begin(), _R1xR2.end(), _R.begin()+jN*_sdim, 0.0) / _norm;
}
template<class TV> static Matrix<T,TV::m+1> affine_register_helper(RawArray<const TV> X0, RawArray<const TV> X1, RawArray<const T> mass) {
typedef typename TV::Scalar T;
static const int d = TV::m;
const int n = X0.size();
GEODE_ASSERT(n==X1.size() && (mass.size()==0 || mass.size()==n));
const bool weighted = mass.size() == n;
const T total = weighted ? mass.sum() : n;
if (!total)
return Matrix<T,d+1>::identity_matrix();
// Compute centers of mass
TV c0, c1;
if (weighted) {
for (int i=0;i<n;i++) {
c0 += mass[i]*X0[i];
c1 += mass[i]*X1[i];
}
} else {
c0 = X0.sum();
c1 = X1.sum();
}
c0 /= total;
c1 /= total;
// Compute covariances
SymmetricMatrix<T,d> cov00 = scaled_outer_product(-total,c0);
Matrix<T,d> cov01 = outer_product(-total*c1,c0);
if (weighted)
for(int i=0;i<n;i++) {
cov00 += scaled_outer_product(mass[i],X0[i]);
cov01 += outer_product(mass[i]*X1[i],X0[i]);
}
else
for (int i=0;i<n;i++) {
cov00 += outer_product(X0[i]);
cov01 += outer_product(X1[i],X0[i]);
}
// Compute transform
const Matrix<T,d> A = cov01*cov00.inverse();
const TV t = c1-A*c0;
auto tA = Matrix<T,d+1>::from_linear(A);
tA.set_translation(t);
return tA;
}
void UnconstrainedLocalSearch::update_B_SR1(Matrix& Bk, const Vector& sk, const Vector& gk, const Vector& gk1) {
Vector r = gk1-gk-Bk*sk;
double tmp = r*sk;
// if tmp=0 => sk =0 and gk=gk1. That means that we have converge
// but due to rounding error in the computation of the stopping criteria,
// we cannot detect it.
// If the norm of the correction is too large we skip the correction
// cout << " [update_B_SR1] r*r/r*sk = "<< (r*r)/tmp<<endl;
if ((tmp!=0)&&( fabs((r*r)/tmp)<=1.e8)) Bk += (1/tmp)*outer_product(r,r);
}
void hyperbolic_reflection(r3vector a, r3transform *out) {
matrix_el_t denom = minkowski_self_inner_product(a);
matrix_el_t s = 2.0 / denom;
r3vector swapped = {a[0]*s, a[1]*s, a[2]*-1*s};
outer_product(a, swapped, out);
*out *= -1;
*out += identity_transform;
}
void
rich_cell::
update_eccentricity_and_radius()
{
//Step1: project the points to the x-y plane
std::vector<vnl_vector_fixed<double, 2> > points_2d;
std::vector<bool> checked(all_points_.size(), false);
for (unsigned int i = 0; i<all_points_.size(); i++) {
if (!checked[i]) {
//record the x-y position
vnl_vector_fixed<double, 2> pt(all_points_[i][0],all_points_[i][1]);
points_2d.push_back( pt );
checked[i] = true;
//scan through the list to see if other points in the remaining
//list contains the same x-y position
for (unsigned int j = i+1; j<all_points_.size(); j++) {
if (all_points_[j][0] == all_points_[i][0]
&& all_points_[j][1] == all_points_[i][1])
checked[j] = true;
}
}
}
//Step2: compute the scatter matrix
//
// compute the center
vnl_vector_fixed<double, 2> center(0.0,0.0);
for (unsigned int i = 0; i<points_2d.size(); i++) {
center += points_2d[i];
}
center /= (double)points_2d.size();
vnl_matrix<double> cov_matrix(2,2,0.0);
std::cout<<"points_2d.size = "<<points_2d.size()<<std::endl;
for (unsigned int i = 0; i<points_2d.size(); i++) {
cov_matrix += outer_product(points_2d[i]-center, points_2d[i]-center);
}
cov_matrix /= (double)points_2d.size();
//perform eigen-value decomposition to get the semi-major (a) and minor (b)
vnl_svd<double> svd_from( cov_matrix );
//eccentricity=sqrt( 1-(b^2/a^2) ). If the shape is close to
//circular, the value is 0.
//eccentricity_ =vcl_sqrt( 1- (svd_from.W(1)*svd_from.W(1))/(svd_from.W(0)*svd_from.W(0)) );
eccentricity_ =vcl_sqrt( 1- vnl_math_abs(svd_from.W(1)/svd_from.W(0)) );
float long_axis_mag = vcl_sqrt(vnl_math_abs(svd_from.W(0)));
float short_axis_mag = vcl_sqrt(vnl_math_abs(svd_from.W(1)));
average_radius_ = 0.5*(long_axis_mag + short_axis_mag);
}
template<class TV> static Frame<TV> rigid_register_helper(RawArray<const TV> X0, RawArray<const TV> X1, RawArray<const T> mass) {
typedef typename TV::Scalar T;
static const int d = TV::m;
const int n = X0.size();
GEODE_ASSERT(n==X1.size() && (mass.size()==0 || mass.size()==n));
const bool weighted = mass.size() == n;
const T total = weighted ? mass.sum() : n;
if (!total)
return Frame<TV>();
// Compute centers of mass
TV c0, c1;
if (weighted) {
for (int i=0;i<n;i++) {
c0 += mass[i]*X0[i];
c1 += mass[i]*X1[i];
}
} else {
c0 = X0.sum();
c1 = X1.sum();
}
c0 /= total;
c1 /= total;
// Compute covariance
Matrix<T,d> cov = outer_product(-total*c1,c0);
if (weighted)
for(int i=0;i<n;i++)
cov += outer_product(mass[i]*X1[i],X0[i]);
else
for (int i=0;i<n;i++)
cov += outer_product(X1[i],X0[i]);
// Compute frame from covariance
Matrix<T,d> U,V;DiagonalMatrix<T,d> D;
cov.fast_singular_value_decomposition(U,D,V);
Rotation<TV> q(U.times_transpose(V));
return Frame<TV>(c1-q*c0,q);
}
vnlMatrixType calcCov(const vnlMatrixType& X,
const vnlMatrixType& Y,
const vnlVectorType& mu_x,
const vnlVectorType& mu_y)
{
vnlMatrixType Sigma_xy(X.rows(), X.rows(), 0.0);
for (unsigned j = 0; j < X.cols(); j++) {
Sigma_xy += outer_product(Y.get_column(j) - mu_y, X.get_column(j) - mu_x);
}
Sigma_xy /= X.cols();
return Sigma_xy;
}
const Matrix rotation( const Vector& axis, const double angle )
{
const double cosa = cos(angle);
const double sina = sin(angle);
Vector ax = axis / sqrt( axis*axis );
Matrix op;
outer_product( ax, ax, op );
Matrix result = unitMatrix() * cosa
+ op * ( 1.0 - cosa )
+ PauliMatrix[0]*axis[0] * sina
+ PauliMatrix[1]*axis[1] * sina
+ PauliMatrix[2]*axis[2] * sina;
return result;
}
int power_eigensolution(const MT& Ap, MT& Q, MT& L, int max_sol)
{
typedef typename MT::VectorType VT;
MT A = Ap;
int n = min(MT::get_v_dim(), max_sol);
for(int i=0;i<n;++i)
{
// Seed the eigenvector estimate
VT q(1);
float l,l_old;
// As long as we haven't reached the max iterations and the
// eigenvalue has not converged, do
int k=0;
for(; k<2 || k<KMAX && (l-l_old > EV_THRESH * l) ; ++k)
{
// Multiply the eigenvector estimate onto A
const VT z = A * q;
// Check that we did not get the null vector.
float z_len = z.length();
if(z_len < EV_THRESH)
return i;
// Normalize to get the new eigenvector
q = z/z_len;
// Record the old eigenvalue estimate and get a new estimate.
l_old = l;
l = dot(q, A * q);
}
// If we hit the max iterations, we also don't trust the eigensolution
if(k==KMAX)
return i;
// Update the solution by adding the eigenvector to Q and
// the eigenvalue to the diagonal of L.
Q[i] = q;
L[i][i] = l;
// Update A by subtracting the subspace represented by the
// eigensolution just found. This is called the method of
// deflation.
MT B;
outer_product(q,q,B);
A = A - l * B;
}
return n;
}
int power_eigensolution(const MT& Ap, MT& Q, MT& L, unsigned int max_sol)
{
L = MT(0);
typedef typename MT::VectorType VT;
MT A = Ap;
unsigned int n = s_min(MT::get_v_dim(), max_sol);
for(unsigned int i=0;i<n;++i)
{
// Seed the eigenvector estimate
VT q(1);
q.normalize();
double l=123,l_old;
// As long as we haven't reached the max iterations and the
// eigenvalue has not converged, do
unsigned int k=0;
do
{
const VT z = A * q;
double z_len = length(z);
if(z_len < EV_THRESH) return i;
l_old = l;
l = dot(q, z)>0 ? z_len : -z_len;
q = z/z_len;
if(++k==KMAX)
return i;
}
while((fabs(l-l_old) > fabs(EV_THRESH * l)) || k<2);
// Update the solution by adding the eigenvector to Q and
// the eigenvalue to the diagonal of L.
Q[i] = q;
L[i][i] = l;
// Update A by subtracting the subspace represented by the
// eigensolution just found. This is called the method of
// deflation.
MT B;
outer_product(q,q,B);
A = A - l * B;
}
return n;
}
SEXP invariantSimilarityHelper(
typename itk::Image< float , ImageDimension >::Pointer image1,
typename itk::Image< float , ImageDimension >::Pointer image2,
SEXP r_thetas, SEXP r_lsits, SEXP r_WM, SEXP r_scale,
SEXP r_doreflection, SEXP r_txfn )
{
unsigned int mibins = 20;
unsigned int localSearchIterations =
Rcpp::as< unsigned int >( r_lsits ) ;
std::string whichMetric = Rcpp::as< std::string >( r_WM );
std::string txfn = Rcpp::as< std::string >( r_txfn );
bool useprincaxis = true;
typedef typename itk::ImageMaskSpatialObject<ImageDimension>::ImageType
maskimagetype;
typename maskimagetype::Pointer mask = ITK_NULLPTR;
Rcpp::NumericVector thetas( r_thetas );
Rcpp::NumericVector vector_r( r_thetas ) ;
Rcpp::IntegerVector dims( 1 );
Rcpp::IntegerVector doReflection( r_doreflection );
unsigned int vecsize = thetas.size();
dims[0]=0;
typedef float PixelType;
typedef double RealType;
RealType bestscale = Rcpp::as< RealType >( r_scale ) ;
typedef itk::Image< PixelType , ImageDimension > ImageType;
if( image1.IsNotNull() & image2.IsNotNull() )
{
typedef typename itk::ImageMomentsCalculator<ImageType> ImageCalculatorType;
typedef itk::AffineTransform<RealType, ImageDimension> AffineType0;
typedef itk::AffineTransform<RealType, ImageDimension> AffineType;
typedef typename ImageCalculatorType::MatrixType MatrixType;
typedef itk::Vector<float, ImageDimension> VectorType;
VectorType ccg1;
VectorType cpm1;
MatrixType cpa1;
VectorType ccg2;
VectorType cpm2;
MatrixType cpa2;
typename ImageCalculatorType::Pointer calculator1 =
ImageCalculatorType::New();
typename ImageCalculatorType::Pointer calculator2 =
ImageCalculatorType::New();
calculator1->SetImage( image1 );
calculator2->SetImage( image2 );
typename ImageCalculatorType::VectorType fixed_center;
fixed_center.Fill(0);
typename ImageCalculatorType::VectorType moving_center;
moving_center.Fill(0);
try
{
calculator1->Compute();
fixed_center = calculator1->GetCenterOfGravity();
ccg1 = calculator1->GetCenterOfGravity();
cpm1 = calculator1->GetPrincipalMoments();
cpa1 = calculator1->GetPrincipalAxes();
try
{
calculator2->Compute();
moving_center = calculator2->GetCenterOfGravity();
ccg2 = calculator2->GetCenterOfGravity();
cpm2 = calculator2->GetPrincipalMoments();
cpa2 = calculator2->GetPrincipalAxes();
}
catch( ... )
{
fixed_center.Fill(0);
}
}
catch( ... )
{
// Rcpp::Rcerr << " zero image1 error ";
}
if ( vnl_math_abs( bestscale - 1.0 ) < 1.e-6 )
{
RealType volelt1 = 1;
RealType volelt2 = 1;
for ( unsigned int d=0; d<ImageDimension; d++)
{
volelt1 *= image1->GetSpacing()[d];
volelt2 *= image2->GetSpacing()[d];
}
bestscale =
( calculator2->GetTotalMass() * volelt2 )/
( calculator1->GetTotalMass() * volelt1 );
RealType powlev = 1.0 / static_cast<RealType>(ImageDimension);
bestscale = vcl_pow( bestscale , powlev );
}
unsigned int eigind1 = 1;
unsigned int eigind2 = 1;
if( ImageDimension == 3 )
{
eigind1 = 2;
}
typedef vnl_vector<RealType> EVectorType;
typedef vnl_matrix<RealType> EMatrixType;
EVectorType evec1_2ndary = cpa1.GetVnlMatrix().get_row( eigind2 );
EVectorType evec1_primary = cpa1.GetVnlMatrix().get_row( eigind1 );
EVectorType evec2_2ndary = cpa2.GetVnlMatrix().get_row( eigind2 );
EVectorType evec2_primary = cpa2.GetVnlMatrix().get_row( eigind1 );
/** Solve Wahba's problem http://en.wikipedia.org/wiki/Wahba%27s_problem */
EMatrixType B = outer_product( evec2_primary, evec1_primary );
if( ImageDimension == 3 )
{
B = outer_product( evec2_2ndary, evec1_2ndary )
+ outer_product( evec2_primary, evec1_primary );
}
vnl_svd<RealType> wahba( B );
vnl_matrix<RealType> A_solution = wahba.V() * wahba.U().transpose();
A_solution = vnl_inverse( A_solution );
RealType det = vnl_determinant( A_solution );
if( ( det < 0 ) )
{
vnl_matrix<RealType> id( A_solution );
id.set_identity();
for( unsigned int i = 0; i < ImageDimension; i++ )
{
if( A_solution( i, i ) < 0 )
{
id( i, i ) = -1.0;
}
}
A_solution = A_solution * id.transpose();
}
if ( doReflection[0] == 1 || doReflection[0] == 3 )
{
vnl_matrix<RealType> id( A_solution );
id.set_identity();
id = id - 2.0 * outer_product( evec2_primary , evec2_primary );
A_solution = A_solution * id;
}
if ( doReflection[0] > 1 )
{
vnl_matrix<RealType> id( A_solution );
id.set_identity();
id = id - 2.0 * outer_product( evec1_primary , evec1_primary );
A_solution = A_solution * id;
}
typename AffineType::Pointer affine1 = AffineType::New();
typename AffineType::OffsetType trans = affine1->GetOffset();
itk::Point<RealType, ImageDimension> trans2;
for( unsigned int i = 0; i < ImageDimension; i++ )
{
trans[i] = moving_center[i] - fixed_center[i];
trans2[i] = fixed_center[i] * ( 1 );
}
affine1->SetIdentity();
affine1->SetOffset( trans );
if( useprincaxis )
{
affine1->SetMatrix( A_solution );
}
affine1->SetCenter( trans2 );
if( ImageDimension > 3 )
{
return EXIT_SUCCESS;
}
vnl_vector<RealType> evec_tert;
if( ImageDimension == 3 )
{ // try to rotate around tertiary and secondary axis
evec_tert = vnl_cross_3d( evec1_primary, evec1_2ndary );
}
if( ImageDimension == 2 )
{ // try to rotate around tertiary and secondary axis
evec_tert = evec1_2ndary;
evec1_2ndary = evec1_primary;
}
itk::Vector<RealType, ImageDimension> axis2;
itk::Vector<RealType, ImageDimension> axis1;
for( unsigned int d = 0; d < ImageDimension; d++ )
{
axis1[d] = evec_tert[d];
axis2[d] = evec1_2ndary[d];
}
typename AffineType::Pointer simmer = AffineType::New();
simmer->SetIdentity();
simmer->SetCenter( trans2 );
simmer->SetOffset( trans );
typename AffineType0::Pointer affinesearch = AffineType0::New();
affinesearch->SetIdentity();
affinesearch->SetCenter( trans2 );
typedef itk::MultiStartOptimizerv4 OptimizerType;
typename OptimizerType::MetricValuesListType metricvalues;
typename OptimizerType::Pointer mstartOptimizer = OptimizerType::New();
typedef itk::CorrelationImageToImageMetricv4
<ImageType, ImageType, ImageType> GCMetricType;
typedef itk::MattesMutualInformationImageToImageMetricv4
<ImageType, ImageType, ImageType> MetricType;
typename MetricType::ParametersType newparams( affine1->GetParameters() );
typename GCMetricType::Pointer gcmetric = GCMetricType::New();
gcmetric->SetFixedImage( image1 );
gcmetric->SetVirtualDomainFromImage( image1 );
gcmetric->SetMovingImage( image2 );
gcmetric->SetMovingTransform( simmer );
gcmetric->SetParameters( newparams );
typename MetricType::Pointer mimetric = MetricType::New();
mimetric->SetNumberOfHistogramBins( mibins );
mimetric->SetFixedImage( image1 );
mimetric->SetMovingImage( image2 );
mimetric->SetMovingTransform( simmer );
mimetric->SetParameters( newparams );
if( mask.IsNotNull() )
{
typename itk::ImageMaskSpatialObject<ImageDimension>::Pointer so =
itk::ImageMaskSpatialObject<ImageDimension>::New();
so->SetImage( const_cast<maskimagetype *>( mask.GetPointer() ) );
mimetric->SetFixedImageMask( so );
gcmetric->SetFixedImageMask( so );
}
typedef itk::ConjugateGradientLineSearchOptimizerv4 LocalOptimizerType;
typename LocalOptimizerType::Pointer localoptimizer =
LocalOptimizerType::New();
RealType localoptimizerlearningrate = 0.1;
localoptimizer->SetLearningRate( localoptimizerlearningrate );
localoptimizer->SetMaximumStepSizeInPhysicalUnits(
localoptimizerlearningrate );
localoptimizer->SetNumberOfIterations( localSearchIterations );
localoptimizer->SetLowerLimit( 0 );
localoptimizer->SetUpperLimit( 2 );
localoptimizer->SetEpsilon( 0.1 );
localoptimizer->SetMaximumLineSearchIterations( 50 );
localoptimizer->SetDoEstimateLearningRateOnce( true );
localoptimizer->SetMinimumConvergenceValue( 1.e-6 );
localoptimizer->SetConvergenceWindowSize( 5 );
if( true )
{
typedef typename MetricType::FixedSampledPointSetType PointSetType;
typedef typename PointSetType::PointType PointType;
typename PointSetType::Pointer pset(PointSetType::New());
unsigned int ind=0;
unsigned int ct=0;
itk::ImageRegionIteratorWithIndex<ImageType> It(image1,
image1->GetLargestPossibleRegion() );
for( It.GoToBegin(); !It.IsAtEnd(); ++It )
{
// take every N^th point
if ( ct % 10 == 0 )
{
PointType pt;
image1->TransformIndexToPhysicalPoint( It.GetIndex(), pt);
pset->SetPoint(ind, pt);
ind++;
}
ct++;
}
mimetric->SetFixedSampledPointSet( pset );
mimetric->SetUseFixedSampledPointSet( true );
gcmetric->SetFixedSampledPointSet( pset );
gcmetric->SetUseFixedSampledPointSet( true );
}
if ( whichMetric.compare("MI") == 0 ) {
mimetric->Initialize();
typedef itk::RegistrationParameterScalesFromPhysicalShift<MetricType>
RegistrationParameterScalesFromPhysicalShiftType;
typename RegistrationParameterScalesFromPhysicalShiftType::Pointer
shiftScaleEstimator =
RegistrationParameterScalesFromPhysicalShiftType::New();
shiftScaleEstimator->SetMetric( mimetric );
shiftScaleEstimator->SetTransformForward( true );
typename RegistrationParameterScalesFromPhysicalShiftType::ScalesType
movingScales( simmer->GetNumberOfParameters() );
shiftScaleEstimator->EstimateScales( movingScales );
mstartOptimizer->SetScales( movingScales );
mstartOptimizer->SetMetric( mimetric );
localoptimizer->SetMetric( mimetric );
localoptimizer->SetScales( movingScales );
}
if ( whichMetric.compare("MI") != 0 ) {
gcmetric->Initialize();
typedef itk::RegistrationParameterScalesFromPhysicalShift<GCMetricType>
RegistrationParameterScalesFromPhysicalShiftType;
typename RegistrationParameterScalesFromPhysicalShiftType::Pointer
shiftScaleEstimator =
RegistrationParameterScalesFromPhysicalShiftType::New();
shiftScaleEstimator->SetMetric( gcmetric );
shiftScaleEstimator->SetTransformForward( true );
typename RegistrationParameterScalesFromPhysicalShiftType::ScalesType
movingScales( simmer->GetNumberOfParameters() );
shiftScaleEstimator->EstimateScales( movingScales );
mstartOptimizer->SetScales( movingScales );
mstartOptimizer->SetMetric( gcmetric );
localoptimizer->SetMetric( gcmetric );
localoptimizer->SetScales( movingScales );
}
typename OptimizerType::ParametersListType parametersList =
mstartOptimizer->GetParametersList();
affinesearch->SetIdentity();
affinesearch->SetCenter( trans2 );
affinesearch->SetOffset( trans );
for ( unsigned int i = 0; i < vecsize; i++ )
{
RealType ang1 = thetas[i];
RealType ang2 = 0; // FIXME should be psi
vector_r[ i ]=0;
if( ImageDimension == 3 )
{
for ( unsigned int jj = 0; jj < vecsize; jj++ )
{
ang2=thetas[jj];
affinesearch->SetIdentity();
affinesearch->SetCenter( trans2 );
affinesearch->SetOffset( trans );
if( useprincaxis )
{
affinesearch->SetMatrix( A_solution );
}
affinesearch->Rotate3D(axis1, ang1, 1);
affinesearch->Rotate3D(axis2, ang2, 1);
affinesearch->Scale( bestscale );
simmer->SetMatrix( affinesearch->GetMatrix() );
parametersList.push_back( simmer->GetParameters() );
}
}
if( ImageDimension == 2 )
{
affinesearch->SetIdentity();
affinesearch->SetCenter( trans2 );
affinesearch->SetOffset( trans );
if( useprincaxis )
{
affinesearch->SetMatrix( A_solution );
}
affinesearch->Rotate2D( ang1, 1);
affinesearch->Scale( bestscale );
simmer->SetMatrix( affinesearch->GetMatrix() );
typename AffineType::ParametersType pp =
simmer->GetParameters();
//pp[1]=ang1;
//pp[0]=bestscale;
parametersList.push_back( simmer->GetParameters() );
}
}
mstartOptimizer->SetParametersList( parametersList );
if( localSearchIterations > 0 )
{
mstartOptimizer->SetLocalOptimizer( localoptimizer );
}
mstartOptimizer->StartOptimization();
typename AffineType::Pointer bestaffine = AffineType::New();
bestaffine->SetCenter( trans2 );
bestaffine->SetParameters( mstartOptimizer->GetBestParameters() );
if ( txfn.length() > 3 )
{
typename AffineType::Pointer bestaffine = AffineType::New();
bestaffine->SetCenter( trans2 );
bestaffine->SetParameters( mstartOptimizer->GetBestParameters() );
typedef itk::TransformFileWriter TransformWriterType;
typename TransformWriterType::Pointer transformWriter =
TransformWriterType::New();
transformWriter->SetInput( bestaffine );
transformWriter->SetFileName( txfn.c_str() );
transformWriter->Update();
}
metricvalues = mstartOptimizer->GetMetricValuesList();
for ( unsigned int k = 0; k < metricvalues.size(); k++ )
{
vector_r[k] = metricvalues[k];
}
dims[0] = vecsize;
vector_r.attr( "dim" ) = vecsize;
return Rcpp::wrap( vector_r );
}
else
{
return Rcpp::wrap( vector_r );
}
}