///////////////////////
///// Inference /////
///////////////////////
void expAndNormalize ( MatrixXf & out, const MatrixXf & in ) {
out.resize( in.rows(), in.cols() );
for( int i=0; i<out.cols(); i++ ){
VectorXf b = in.col(i);
b.array() -= b.maxCoeff();
b = b.array().exp();
out.col(i) = b / b.array().sum();
}
}
float rttemp1( VectorXf t_reg, VectorXf curvei, VectorXf curvek,int nknots, float lambda, VectorXf initial){
//Calculate the cost of the given warping
// t_reg : time grid of y_reg
// curvei : query curve
// curvek : reference curve
// nknots : number of knots
// lambda : time distortion penalty parameter
// initial: position of knots on the simplex
int N = t_reg.size();
VectorXf struct_ = VectorXf::LinSpaced( nknots+2, t_reg(0) , t_reg.maxCoeff() );
VectorXf hik(N);
VectorXf Q(2+ initial.size()) ; //Solution with the placement of the knots on the simplex
Q(0) = t_reg(0) ; Q(1+ initial.size()) = t_reg.maxCoeff() ; Q.segment(1, initial.size()) = initial;
hik = fnvalspapi( struct_, Q , t_reg); // compute the new internal time-scale
//cout << "hik: " << hik.transpose() << endl;
//cout << "Monotonicity Checked on Hik: " << MonotonicityCheck(hik) << endl;
//if( MonotonicityCheck(hik) == -1) { cout <<" Q.transpose() is :"<< Q.transpose() << endl;}
return ( (fnvalspapi(t_reg,curvei,hik)-fnvalspapi(t_reg,curvek ,t_reg)).array().pow(2).sum() + lambda * (hik - t_reg).array().pow(2).sum() );
}
rthink_Output rthik_SA(VectorXf t_reg, VectorXf curvei, VectorXf curvek,int nknots, float lambda){
// Random Search solver for the optimization problem of pairwise warping
// t_reg : time_grid
// curvei: query curve
// curvek: reference curve
// nknots : number of knots
// lambda : time distortion penalty parameter
int k=0; float OldSol, bk;
VectorXf xk(nknots);
bk = (t_reg.maxCoeff() - t_reg(0))/(1+ float(nknots )); //Distance between adjacent knots and edges-knots
xk = VectorXf::LinSpaced(nknots , t_reg(0) + bk , - bk + t_reg.maxCoeff() ); //Initial candidate solution with equispaced knots
VectorXf xn(nknots); VectorXf help(2+nknots);
float NewSol;
OldSol = rttemp1(t_reg, curvei, curvek, nknots , lambda, xk); //Cost of initial solution
int z= 99*nknots; //Number of random search to do (proportional to the # of knots)
//srand(1); //Fix the seed to have reproducable behaviour
VectorXf Steps(z); Steps = (ArrayXf::Random(z)+1.)/2.; //Generate possible random pertubations magnitude
VectorXi Posit(z); for (int u=0; u <z; u++ ) Posit(u) =1+ rand()%(nknots+0); //Generate list of positions to purturb
k=0;
while((OldSol > .0001) && (k<z)) {
xn = NewSolution(xk,Steps(k), Posit(k), t_reg ); //Get a new solution
NewSol = rttemp1(t_reg, curvei, curvek, nknots, lambda, xn); //Cost of new solution
if ( (NewSol < OldSol) ) { //If it's better than the old one, use it.
OldSol= NewSol; xk= xn;
}
k++;
}
VectorXf x3 = VectorXf::LinSpaced(2+nknots, t_reg(0), t_reg( t_reg.size()-1));
help(0) = t_reg(0) ; help(nknots+1) = t_reg( t_reg.size()-1) ; help.segment(1,nknots) = xk;
rthink_Output G;
G.Val = OldSol;
G.Mapping = fnvalspapi( x3 , help , t_reg);
return G ;
}
VectorXf NewSolution( VectorXf x0 , float Step, int point, VectorXf t_reg){
//generate new solution on the simplex defined in [t_reg(0)-x0-t_reg(N-1)] using a displacement of size Step
// x0 : initial solution
// Step : displacement size
// point: knot to perturb
// t_reg: time_grid
VectorXf InitConf (2+ x0.size());
InitConf(0) = t_reg(0); InitConf( x0.size()+1) = t_reg.maxCoeff() ; InitConf.segment(1, x0.size()) = x0;
float LowBou = InitConf(point-1);
float UppBou = InitConf(point+1);
float New_State = (UppBou - LowBou) * Step + LowBou;
InitConf(point) = New_State;
//if ( MonotonicityCheck( InitConf.segment(1, x0.size()) ) == -1) {
// cout << "We generated a unacceptable solutiion" << endl << "Initial seed was : " << x0.transpose()
// << endl << " and we produced :"<<InitConf.segment(1, x0.size()).transpose() << endl;}
return (InitConf.segment(1, x0.size()) ) ;
}
void wavePlotter::autoScaleRange(VectorXf &data)
{
lowRange = data.minCoeff();//Utils::ofMin(&data(0), DATA_SIZE);
highRange = data.maxCoeff();//Utils::ofMax(&data(0), DATA_SIZE);
}
void D3DCloudProjector::projectCloud(int id, const sensor_msgs::PointCloud& data, const std::vector<int>& interest_region_indices) {
MatrixXf& oriented = orienter_->oriented_clouds_[id];
// -- Get a copy of the projected points.
MatrixXf projected(oriented.rows(), 2);
int c=0;
for(int i=0; i<3; ++i) {
if(i == axis_of_projection_)
continue;
projected.col(c) = oriented.col(i);
++c;
}
// -- Transform into pixel units. projected is currently in meters, centered at 0.
//projected *= pixels_per_meter_;
for(int i=0; i<projected.rows(); ++i) {
projected(i, 0) *= pixels_per_meter_;
projected(i, 1) *= pixels_per_meter_;
}
// -- Find min and max of u and v. TODO: noise sensitivity?
// u is the col number in the image plane, v is the row number.
float min_v = FLT_MAX;
float min_u = FLT_MAX;
float max_v = -FLT_MAX;
float max_u = -FLT_MAX;
for(int i=0; i<projected.rows(); ++i) {
float u = projected(i, 0);
float v = projected(i, 1);
if(u < min_u)
min_u = u;
if(u > max_u)
max_u = u;
if(v < min_v)
min_v = v;
if(v > max_v)
max_v = v;
}
// -- Translate to coordinate system where (0,0) is the upper right of the image.
for(int i=0; i<projected.rows(); ++i) {
projected(i, 0) -= min_u;
projected(i, 1) = max_v - projected(i, 1);
}
// -- Get the max depth.
float max_depth = -FLT_MAX;
float min_depth = FLT_MAX;
for(int i=0; i<oriented.rows(); ++i) {
if(oriented(i, axis_of_projection_) > max_depth)
max_depth = oriented(i, axis_of_projection_);
if(oriented(i, axis_of_projection_) < min_depth)
min_depth = oriented(i, axis_of_projection_);
}
// -- Compute the normalized depths. Depths are between 0 and 1, with 1 meaning closest and 0 meaning furthest.
VectorXf depths = oriented.col(axis_of_projection_);
if(axis_of_projection_ == 1)
depths = -depths;
depths = depths.cwise() - depths.minCoeff();
depths = depths / depths.maxCoeff();
// -- Fill the IplImages.
assert(sizeof(float) == 4);
CvSize size = cvSize(ceil(max_u - min_u), ceil(max_v - min_v));
IplImage* acc = cvCreateImage(size, IPL_DEPTH_32F, 1);
IplImage* intensity = cvCreateImage(size, IPL_DEPTH_32F, 1);
IplImage* depth = cvCreateImage(size, IPL_DEPTH_32F, 1);
cvSetZero(acc);
cvSetZero(depth);
cvSetZero(intensity);
assert(projected.rows() == (int)interest_region_indices.size());
for(int i=0; i<projected.rows(); ++i) {
int row = floor(projected(i, 1));
int col = floor(projected(i, 0));
// Update accumulator.
assert(interest_region_indices[i] < (int)data.channels[0].values.size() && (int)interest_region_indices[i] >= 0);
((float*)(acc->imageData + row * acc->widthStep))[col]++;
// Add to intensity values.
float val = (float)data.channels[0].values[interest_region_indices[i]] / 255.0 * (3.0 / 4.0) + 0.25;
assert(val <= 1.0 && val >= 0.0);
((float*)(intensity->imageData + row * intensity->widthStep))[col] += val;
// Add to depth values.
((float*)(depth->imageData + row * depth->widthStep))[col] += depths(i); //
}
// -- Normalize by the number of points falling in each pixel.
for(int v=0; v<acc->height; ++v) {
float* intensity_ptr = (float*)(intensity->imageData + v * intensity->widthStep);
float* depth_ptr = (float*)(depth->imageData + v * depth->widthStep);
float* acc_ptr = (float*)(acc->imageData + v * acc->widthStep);
for(int u=0; u<acc->width; ++u) {
if(*acc_ptr == 0) {
*intensity_ptr = 0;
*depth_ptr = 0;
}
else {
*intensity_ptr = *intensity_ptr / *acc_ptr;
*depth_ptr = *depth_ptr / *acc_ptr;
}
intensity_ptr++;
depth_ptr++;
acc_ptr++;
}
}
// -- Store images.
depth_projections_.push_back(depth);
intensity_projections_.push_back(intensity);
// -- Debugging.
if(debug_) {
float scale = 10;
IplImage* intensity_big = cvCreateImage(cvSize(((float)intensity->width)*scale, ((float)intensity->height)*scale), intensity->depth, intensity->nChannels);
cvResize(intensity, intensity_big, CV_INTER_AREA);
IplImage* depth_big = cvCreateImage(cvSize(((float)depth->width)*scale, ((float)depth->height)*scale), depth->depth, depth->nChannels);
cvResize(depth, depth_big, CV_INTER_AREA);
CVSHOW("Intensity Image", intensity_big);
CVSHOW("Depth Image", depth_big);
cvWaitKey(0);
cvDestroyWindow("Intensity Image");
cvDestroyWindow("Depth Image");
}
// -- Clean up.
cvReleaseImage(&acc);
}
void CloudProjector::_compute() {
assert(orienter_);
assert(orienter_->getOutputCloud());
assert(!depth_projection_);
assert(!intensity_projection_);
MatrixXf& oriented = *orienter_->getOutputCloud();
VectorXf& intensities = *orienter_->getOutputIntensity();
// -- Get a copy of the projected points.
MatrixXf projected(oriented.rows(), 2);
int c=0;
for(int i=0; i<3; ++i) {
if(i == axis_of_projection_)
continue;
projected.col(c) = oriented.col(i);
++c;
}
// -- Transform into pixel units. projected is currently in meters, centered at 0.
//projected *= pixels_per_meter_;
for(int i=0; i<projected.rows(); ++i) {
projected(i, 0) *= pixels_per_meter_;
projected(i, 1) *= pixels_per_meter_;
}
// -- Find min and max of u and v. TODO: noise sensitivity?
// u is the col number in the image plane, v is the row number.
float min_v = FLT_MAX;
float min_u = FLT_MAX;
float max_v = -FLT_MAX;
float max_u = -FLT_MAX;
for(int i=0; i<projected.rows(); ++i) {
float u = projected(i, 0);
float v = projected(i, 1);
if(u < min_u)
min_u = u;
if(u > max_u)
max_u = u;
if(v < min_v)
min_v = v;
if(v > max_v)
max_v = v;
}
// -- Translate to coordinate system where (0,0) is the upper right of the image.
for(int i=0; i<projected.rows(); ++i) {
projected(i, 0) -= min_u;
projected(i, 1) = max_v - projected(i, 1);
}
// -- Get the max depth.
float max_depth = -FLT_MAX;
float min_depth = FLT_MAX;
for(int i=0; i<oriented.rows(); ++i) {
if(oriented(i, axis_of_projection_) > max_depth)
max_depth = oriented(i, axis_of_projection_);
if(oriented(i, axis_of_projection_) < min_depth)
min_depth = oriented(i, axis_of_projection_);
}
// -- Compute the normalized depths. Depths are between 0 and 1, with 1 meaning closest and 0 meaning furthest.
VectorXf depths = oriented.col(axis_of_projection_);
if(axis_of_projection_ == 1)
depths = -depths;
depths = (depths.array() - depths.minCoeff()).matrix();
depths = depths / depths.maxCoeff();
// -- Fill the IplImages.
assert(sizeof(float) == 4);
CvSize size = cvSize(ceil(max_u - min_u) + 1, ceil(max_v - min_v) + 1);
float pad_width = 0;
if(min_width_ > 0 && size.width < min_width_) {
pad_width = (float)(min_width_ - size.width) / 2.;
size.width = min_width_;
}
float pad_height = 0;
if(min_height_ > 0 && size.height < min_height_) {
pad_height = (float)(min_height_ - size.height) / 2.;
size.height = min_height_;
}
IplImage* acc = cvCreateImage(size, IPL_DEPTH_32F, 1);
intensity_projection_ = cvCreateImage(size, IPL_DEPTH_32F, 1);
depth_projection_ = cvCreateImage(size, IPL_DEPTH_32F, 1);
cvSetZero(acc);
cvSetZero(depth_projection_);
cvSetZero(intensity_projection_);
assert(intensities.rows() == projected.rows());
for(int i=0; i<projected.rows(); ++i) {
int row = floor(projected(i, 1) + pad_height);
int col = floor(projected(i, 0) + pad_width);
assert(row < size.height && row >= 0 && col < size.width && col >= 0);
// Update accumulator.
((float*)(acc->imageData + row * acc->widthStep))[col]++;
// Update intensity values.
float val = intensities(i) / 255.0 * (3.0 / 4.0) + 0.25;
assert(val <= 1.0 && val >= 0.0);
((float*)(intensity_projection_->imageData + row * intensity_projection_->widthStep))[col] += val;
// Update depth values.
((float*)(depth_projection_->imageData + row * depth_projection_->widthStep))[col] += depths(i); //
}
// -- Normalize by the number of points falling in each pixel.
for(int v=0; v<acc->height; ++v) {
float* intensity_ptr = (float*)(intensity_projection_->imageData + v * intensity_projection_->widthStep);
float* depth_ptr = (float*)(depth_projection_->imageData + v * depth_projection_->widthStep);
float* acc_ptr = (float*)(acc->imageData + v * acc->widthStep);
for(int u=0; u<acc->width; ++u) {
if(*acc_ptr == 0) {
*intensity_ptr = 0;
*depth_ptr = 0;
}
else {
*intensity_ptr = *intensity_ptr / *acc_ptr;
*depth_ptr = *depth_ptr / *acc_ptr;
}
intensity_ptr++;
depth_ptr++;
acc_ptr++;
}
}
// -- Blur the images. TODO: depth too?
cvSmooth(intensity_projection_, intensity_projection_, CV_GAUSSIAN, smoothing_, smoothing_);
// -- Clean up.
cvReleaseImage(&acc);
}
