Base ceres scan matcher on correspondence cost function (#1085)
- Base ceres scan matcher on correspondence cost function instead of probabilities - Step towards [RFC 0019](https://github.com/googlecartographer/rfcs/blob/master/text/0019-probability-grid-and-submap2d-restructuring.md)master
parent
9b3606b7c0
commit
f4937b5cc6
|
@ -60,8 +60,9 @@ void Grid2D::SetCorrespondenceCost(const Eigen::Array2i& cell_index,
|
|||
|
||||
// Returns the probability of the cell with 'cell_index'.
|
||||
float Grid2D::GetCorrespondenceCost(const Eigen::Array2i& cell_index) const {
|
||||
LOG(FATAL) << "Not implemented";
|
||||
return kUnknownCorrespondenceValue;
|
||||
if (!limits().Contains(cell_index)) return kMaxCorrespondenceCost;
|
||||
return ValueToCorrespondenceCost(
|
||||
correspondence_cost_cells()[ToFlatIndex(cell_index)]);
|
||||
}
|
||||
|
||||
// Returns true if the probability at the specified index is known.
|
||||
|
|
|
@ -22,7 +22,7 @@
|
|||
#include "Eigen/Core"
|
||||
#include "cartographer/common/ceres_solver_options.h"
|
||||
#include "cartographer/common/lua_parameter_dictionary.h"
|
||||
#include "cartographer/mapping/2d/probability_grid.h"
|
||||
#include "cartographer/mapping/2d/grid_2d.h"
|
||||
#include "cartographer/mapping/internal/2d/scan_matching/occupied_space_cost_function_2d.h"
|
||||
#include "cartographer/mapping/internal/2d/scan_matching/rotation_delta_cost_functor_2d.h"
|
||||
#include "cartographer/mapping/internal/2d/scan_matching/translation_delta_cost_functor_2d.h"
|
||||
|
@ -62,7 +62,7 @@ CeresScanMatcher2D::~CeresScanMatcher2D() {}
|
|||
void CeresScanMatcher2D::Match(const Eigen::Vector2d& target_translation,
|
||||
const transform::Rigid2d& initial_pose_estimate,
|
||||
const sensor::PointCloud& point_cloud,
|
||||
const ProbabilityGrid& probability_grid,
|
||||
const Grid2D& grid,
|
||||
transform::Rigid2d* const pose_estimate,
|
||||
ceres::Solver::Summary* const summary) const {
|
||||
double ceres_pose_estimate[3] = {initial_pose_estimate.translation().x(),
|
||||
|
@ -74,7 +74,7 @@ void CeresScanMatcher2D::Match(const Eigen::Vector2d& target_translation,
|
|||
OccupiedSpaceCostFunction2D::CreateAutoDiffCostFunction(
|
||||
options_.occupied_space_weight() /
|
||||
std::sqrt(static_cast<double>(point_cloud.size())),
|
||||
point_cloud, probability_grid),
|
||||
point_cloud, grid),
|
||||
nullptr /* loss function */, ceres_pose_estimate);
|
||||
CHECK_GT(options_.translation_weight(), 0.);
|
||||
problem.AddResidualBlock(
|
||||
|
|
|
@ -22,7 +22,7 @@
|
|||
|
||||
#include "Eigen/Core"
|
||||
#include "cartographer/common/lua_parameter_dictionary.h"
|
||||
#include "cartographer/mapping/2d/probability_grid.h"
|
||||
#include "cartographer/mapping/2d/grid_2d.h"
|
||||
#include "cartographer/mapping/2d/scan_matching/proto/ceres_scan_matcher_options_2d.pb.h"
|
||||
#include "cartographer/sensor/point_cloud.h"
|
||||
#include "ceres/ceres.h"
|
||||
|
@ -48,8 +48,7 @@ class CeresScanMatcher2D {
|
|||
// 'summary'.
|
||||
void Match(const Eigen::Vector2d& target_translation,
|
||||
const transform::Rigid2d& initial_pose_estimate,
|
||||
const sensor::PointCloud& point_cloud,
|
||||
const ProbabilityGrid& probability_grid,
|
||||
const sensor::PointCloud& point_cloud, const Grid2D& grid,
|
||||
transform::Rigid2d* pose_estimate,
|
||||
ceres::Solver::Summary* summary) const;
|
||||
|
||||
|
|
|
@ -19,7 +19,7 @@
|
|||
|
||||
#include "Eigen/Core"
|
||||
#include "Eigen/Geometry"
|
||||
#include "cartographer/mapping/2d/probability_grid.h"
|
||||
#include "cartographer/mapping/2d/grid_2d.h"
|
||||
#include "cartographer/mapping/probability_values.h"
|
||||
#include "cartographer/sensor/point_cloud.h"
|
||||
#include "ceres/ceres.h"
|
||||
|
@ -29,19 +29,18 @@ namespace cartographer {
|
|||
namespace mapping {
|
||||
namespace scan_matching {
|
||||
|
||||
// Computes a cost for matching the 'point_cloud' to the 'probability_grid' with
|
||||
// a 'pose'. The cost increases when points fall into less occupied space, i.e.
|
||||
// at pixels with lower values.
|
||||
// Computes a cost for matching the 'point_cloud' to the 'grid' with
|
||||
// a 'pose'. The cost increases with poorer correspondence of the grid and the
|
||||
// point observation (e.g. points falling into less occupied space).
|
||||
class OccupiedSpaceCostFunction2D {
|
||||
public:
|
||||
static ceres::CostFunction* CreateAutoDiffCostFunction(
|
||||
const double scaling_factor, const sensor::PointCloud& point_cloud,
|
||||
const ProbabilityGrid& probability_grid) {
|
||||
const Grid2D& grid) {
|
||||
return new ceres::AutoDiffCostFunction<OccupiedSpaceCostFunction2D,
|
||||
ceres::DYNAMIC /* residuals */,
|
||||
3 /* pose variables */>(
|
||||
new OccupiedSpaceCostFunction2D(scaling_factor, point_cloud,
|
||||
probability_grid),
|
||||
new OccupiedSpaceCostFunction2D(scaling_factor, point_cloud, grid),
|
||||
point_cloud.size());
|
||||
}
|
||||
|
||||
|
@ -53,9 +52,9 @@ class OccupiedSpaceCostFunction2D {
|
|||
Eigen::Matrix<T, 3, 3> transform;
|
||||
transform << rotation_matrix, translation, T(0.), T(0.), T(1.);
|
||||
|
||||
const GridArrayAdapter adapter(probability_grid_);
|
||||
const GridArrayAdapter adapter(grid_);
|
||||
ceres::BiCubicInterpolator<GridArrayAdapter> interpolator(adapter);
|
||||
const MapLimits& limits = probability_grid_.limits();
|
||||
const MapLimits& limits = grid_.limits();
|
||||
|
||||
for (size_t i = 0; i < point_cloud_.size(); ++i) {
|
||||
// Note that this is a 2D point. The third component is a scaling factor.
|
||||
|
@ -68,7 +67,7 @@ class OccupiedSpaceCostFunction2D {
|
|||
(limits.max().y() - world[1]) / limits.resolution() - 0.5 +
|
||||
static_cast<double>(kPadding),
|
||||
&residual[i]);
|
||||
residual[i] = scaling_factor_ * (1. - residual[i]);
|
||||
residual[i] = scaling_factor_ * residual[i];
|
||||
}
|
||||
return true;
|
||||
}
|
||||
|
@ -79,39 +78,36 @@ class OccupiedSpaceCostFunction2D {
|
|||
public:
|
||||
enum { DATA_DIMENSION = 1 };
|
||||
|
||||
explicit GridArrayAdapter(const ProbabilityGrid& probability_grid)
|
||||
: probability_grid_(probability_grid) {}
|
||||
explicit GridArrayAdapter(const Grid2D& grid) : grid_(grid) {}
|
||||
|
||||
void GetValue(const int row, const int column, double* const value) const {
|
||||
if (row < kPadding || column < kPadding || row >= NumRows() - kPadding ||
|
||||
column >= NumCols() - kPadding) {
|
||||
*value = kMinProbability;
|
||||
*value = kMaxCorrespondenceCost;
|
||||
} else {
|
||||
*value = static_cast<double>(probability_grid_.GetProbability(
|
||||
*value = static_cast<double>(grid_.GetCorrespondenceCost(
|
||||
Eigen::Array2i(column - kPadding, row - kPadding)));
|
||||
}
|
||||
}
|
||||
|
||||
int NumRows() const {
|
||||
return probability_grid_.limits().cell_limits().num_y_cells +
|
||||
2 * kPadding;
|
||||
return grid_.limits().cell_limits().num_y_cells + 2 * kPadding;
|
||||
}
|
||||
|
||||
int NumCols() const {
|
||||
return probability_grid_.limits().cell_limits().num_x_cells +
|
||||
2 * kPadding;
|
||||
return grid_.limits().cell_limits().num_x_cells + 2 * kPadding;
|
||||
}
|
||||
|
||||
private:
|
||||
const ProbabilityGrid& probability_grid_;
|
||||
const Grid2D& grid_;
|
||||
};
|
||||
|
||||
OccupiedSpaceCostFunction2D(const double scaling_factor,
|
||||
const sensor::PointCloud& point_cloud,
|
||||
const ProbabilityGrid& probability_grid)
|
||||
const Grid2D& grid)
|
||||
: scaling_factor_(scaling_factor),
|
||||
point_cloud_(point_cloud),
|
||||
probability_grid_(probability_grid) {}
|
||||
grid_(grid) {}
|
||||
|
||||
OccupiedSpaceCostFunction2D(const OccupiedSpaceCostFunction2D&) = delete;
|
||||
OccupiedSpaceCostFunction2D& operator=(const OccupiedSpaceCostFunction2D&) =
|
||||
|
@ -119,7 +115,7 @@ class OccupiedSpaceCostFunction2D {
|
|||
|
||||
const double scaling_factor_;
|
||||
const sensor::PointCloud& point_cloud_;
|
||||
const ProbabilityGrid& probability_grid_;
|
||||
const Grid2D& grid_;
|
||||
};
|
||||
|
||||
} // namespace scan_matching
|
||||
|
|
Loading…
Reference in New Issue