cv::large_kinfu::VolumeParams Struct Reference

#include <opencv2/rgbd/large_kinfu.hpp>

Collaboration diagram for cv::large_kinfu::VolumeParams:

Static Public Member Functions
static Ptr< VolumeParams >	coarseParams (VolumeType volumeType)
	Coarse set of parameters that provides relatively higher performance at the cost of reconstrution quality.
static Ptr< VolumeParams >	defaultParams (VolumeType volumeType)
	Default set of parameters that provide higher quality reconstruction at the cost of slow performance.

Public Attributes
float	depthTruncThreshold = 0.f
	Threshold for depth truncation in meters.
VolumeType	kind = VolumeType::TSDF
	Kind of Volume.
int	maxWeight = 64
	Max number of frames to integrate per voxel.
Matx44f	pose = Affine3f().translate(Vec3f(-volumSize / 2.f, -volumSize / 2.f, 0.5f)).matrix
	Initial pose of the volume in meters, should be 4x4 float or double matrix.
float	raycastStepFactor = 0.25f
	Length of single raycast step.
int	resolutionX = 128
	Resolution of voxel space.
int	resolutionY = 128
int	resolutionZ = 128
float	tsdfTruncDist = 7.f * voxelSize
	TSDF truncation distance.
int	unitResolution = 0
	Resolution of volumeUnit in voxel space.
float	volumSize = 3.f
	Size of volume in meters.
float	voxelSize = volumSize / 512.f
	Length of voxels in meters.

Member Function Documentation

coarseParams()

static Ptr< VolumeParams > cv::large_kinfu::VolumeParams::coarseParams ( VolumeType volumeType )

static

Coarse set of parameters that provides relatively higher performance at the cost of reconstrution quality.

defaultParams()

static Ptr< VolumeParams > cv::large_kinfu::VolumeParams::defaultParams ( VolumeType volumeType )

static

Default set of parameters that provide higher quality reconstruction at the cost of slow performance.

Member Data Documentation

depthTruncThreshold

float cv::large_kinfu::VolumeParams::depthTruncThreshold = 0.f

Threshold for depth truncation in meters.

Truncates the depth greater than threshold to 0

kind

VolumeType cv::large_kinfu::VolumeParams::kind = VolumeType::TSDF

Kind of Volume.

Values can be TSDF (single volume) or HASHTSDF (hashtable of volume units)

maxWeight

int cv::large_kinfu::VolumeParams::maxWeight = 64

Max number of frames to integrate per voxel.

Represents the max number of frames over which a running average of the TSDF is calculated for a voxel

pose

Matx44f cv::large_kinfu::VolumeParams::pose = Affine3f().translate(Vec3f(-volumSize / 2.f, -volumSize / 2.f, 0.5f)).matrix

Initial pose of the volume in meters, should be 4x4 float or double matrix.

raycastStepFactor

float cv::large_kinfu::VolumeParams::raycastStepFactor = 0.25f

Length of single raycast step.

Describes the percentage of voxel length that is skipped per march

resolutionX

int cv::large_kinfu::VolumeParams::resolutionX = 128

Resolution of voxel space.

Number of voxels in each dimension. Applicable only for TSDF Volume. HashTSDF volume only supports equal resolution in all three dimensions

resolutionY

int cv::large_kinfu::VolumeParams::resolutionY = 128

resolutionZ

int cv::large_kinfu::VolumeParams::resolutionZ = 128

tsdfTruncDist

float cv::large_kinfu::VolumeParams::tsdfTruncDist = 7.f * voxelSize

TSDF truncation distance.

Distances greater than value from surface will be truncated to 1.0

unitResolution

int cv::large_kinfu::VolumeParams::unitResolution = 0

Resolution of volumeUnit in voxel space.

Number of voxels in each dimension for volumeUnit Applicable only for hashTSDF.

volumSize

float cv::large_kinfu::VolumeParams::volumSize = 3.f

Size of volume in meters.

voxelSize

float cv::large_kinfu::VolumeParams::voxelSize = volumSize / 512.f

Length of voxels in meters.

---

The documentation for this struct was generated from the following file:

• opencv2/rgbd/large_kinfu.hpp