![]() |
Tesseract 0.28.4
|
Tesseract Signed Distance Field Geometry. More...
#include <tesseract/common/macros.h>#include <Eigen/Core>#include <Eigen/Geometry>#include <cstdint>#include <functional>#include <memory>#include <vector>#include <tesseract/common/fwd.h>#include <tesseract/geometry/geometry.h>Classes | |
| class | tesseract::geometry::SignedDistanceField |
| A volumetric signed distance field geometry. More... | |
Typedefs | |
| using | tesseract::geometry::SignedDistanceFunction = std::function<double(const Eigen::Vector3d&)> |
| Evaluate the signed distance at a single point in the field's local frame. | |
| using | tesseract::geometry::BatchedSignedDistanceFunction = std::function<std::vector<double>(const std::vector<Eigen::Vector3d>&)> |
| Evaluate the signed distance at many points at once. | |
Enumerations | |
| enum class | tesseract::geometry::SignedDistanceFieldInterpolation : std::uint8_t { TRILINEAR , TRICUBIC } |
| How a grid-backed or sparse field is read between its samples. More... | |
Functions | |
| template<class Archive> | |
| void | tesseract::geometry::serialize (Archive &ar, SignedDistanceField &obj) |
| Format version 1, the first with a version: the interpolation settings, then the band clamp of each NanoVDB grid and the grids as one NanoVDB file, as writeSignedDistanceFieldNVDB writes it. A sparse field has one grid and a multi-level field one per level, finest first; their domain and dimensions follow from the grids, and a field with one level loads as a plain sparse field. A field without grids is dense and stores its sample grid instead. Then the scale. A later format bumps SIGNED_DISTANCE_FIELD_ARCHIVE_VERSION and branches on it here. | |
Tesseract Signed Distance Field Geometry.
| using tesseract::geometry::BatchedSignedDistanceFunction = std::function<std::vector<double>(const std::vector<Eigen::Vector3d>&)> |
Evaluate the signed distance at many points at once.
Receives every sample point in one call and must return one distance per point, in the same order. Use this overload to vectorize / offload the evaluation (e.g. a batched nvblox ESDF GPU query) instead of paying a per-point call.
|
strong |
How a grid-backed or sparse field is read between its samples.
Both pass through the samples and reproduce a linear field exactly. They differ in how many samples a read uses and whether the gradient is continuous across cell faces.
| void tesseract::geometry::serialize | ( | Archive & | ar, |
| SignedDistanceField & | obj ) |
Format version 1, the first with a version: the interpolation settings, then the band clamp of each NanoVDB grid and the grids as one NanoVDB file, as writeSignedDistanceFieldNVDB writes it. A sparse field has one grid and a multi-level field one per level, finest first; their domain and dimensions follow from the grids, and a field with one level loads as a plain sparse field. A field without grids is dense and stores its sample grid instead. Then the scale. A later format bumps SIGNED_DISTANCE_FIELD_ARCHIVE_VERSION and branches on it here.