Pointclouds and PCL Types

struct PointCloud{T<:Caesar._PCL.PointT, P, R}

Convert a PCLPointCloud2 binary data blob into a Caesar._PCL.PointCloud{T} object using a field_map::Caesar._PCL.MsgFieldMap.

Use PointCloud(::Caesar._PCL.PCLPointCloud2) directly or create you own MsgFieldMap:

field_map = Caesar._PCL.createMapping(msg.fields, field_map)

Notes

• Tested on Radar data with height z=constant for all points – i.e. 2D sweeping scan where .height=1.

DevNotes

• TODO .PCLPointCloud2 convert not tested on regular 3D data from structured light or lidar yet, but current implementation should be close (or already working).

References

• https://pointclouds.org/documentation/classpcl11pointcloud.html
• (seems older) https://docs.ros.org/en/hydro/api/pcl/html/conversions8hsource.html#l00123

struct PointXYZ{C<:Colorant, T<:Number} <: Caesar._PCL.PointT

Immutable PointXYZ with color information. E.g. PointXYZ{RGB}, PointXYZ{Gray}, etc.

Aliases

• PointXYZRGB
• PointXYZRGBA

See

• https://pointclouds.org/documentation/structpcl11pointxyz.html
• https://pointclouds.org/documentation/point__types8hppsource.html

struct Header

Immutable Header.

See https://pointclouds.org/documentation/structpcl11pclheader.html

struct PointField

How a point is stored in memory.

• https://pointclouds.org/documentation/structpcl11pclpoint_field.html

struct FieldMapper{T<:Caesar._PCL.PointT}

Which field values to store and how to map them to values during serialization.

• https://docs.ros.org/en/hydro/api/pcl/html/conversions8hsource.html#l00091

struct PCLPointCloud2

Immutable point cloud type. Immutable for performance, computations are more frequent and intensive than anticipated frequency of constructing new clouds.

References:

• https://pointclouds.org/documentation/structpcl11pclpoint_cloud2.html
• https://pointclouds.org/documentation/classpcl11pointcloud.html
• https://pointclouds.org/documentation/common2include2pcl2point__cloud8h_source.html

See also: Caesar._PCL.toROSPointCloud2

ScatterAlign{P,H1,H2} where  {H1 <: Union{<:ManifoldKernelDensity, <:HeatmapGridDensity}, 
                              H2 <: Union{<:ManifoldKernelDensity, <:HeatmapGridDensity}}

Alignment factor between point cloud populations, using either

• a continuous density function cost: ApproxManifoldProducts.mmd, or
• a conventional iterative closest point (ICP) algorithm (when .sample_count < 0).

This factor can support very large density clouds, with sample_count subsampling for individual alignments.

Keyword Options:

• sample_count::Int = 100, number of subsamples to use during each alignment in getSample.
  • Values greater than 0 use MMD alignment, while values less than 0 use ICP alignment.
• bw::Real, the bandwidth to use for mmd distance
• rescale::Real
• N::Int
• cvt::Function, convert function for image when using HeatmapGridDensity.
• useStashing::Bool = false, to switch serialization strategy to using Stashing.
• dataEntry_cloud1::AbstractString = "", blob identifier used with stashing.
• dataEntry_cloud2::AbstractString = "", blob identifier used with stashing.
• dataStoreHint::AbstractString = ""

Example

arp2 = ScatterAlignPose2(img1, img2, 2) # e.g. 2 meters/pixel 

Notes

• Supports two belief "clouds" as either
  • ManifoldKernelDensitys, or
  • HeatmapGridDensitys.
• Stanard cvt argument is lambda function to convert incoming images to user convention of image axes,
  • Geography map default cvt flips image rows so that Pose2 +xy-axes corresponds to img[-x,+y]
    • i.e. rows down is "North" and columns across from top left corner is "East".
• Use rescale to resize the incoming images for lower resolution (faster) correlations
• Both images passed to the construct must have the same type some matrix of type T.
• Experimental support for Stashing based serialization.

DevNotes:

• TODO Upgrade to use other information during alignment process, e.g. point normals for Pose3.

See also: ScatterAlignPose2, ScatterAlignPose3, overlayScanMatcher, Caesar._PCL.alignICP_Simple.

ScatterAlignPose2(im1::Matrix, im2::Matrix, domain; options...)
ScatterAlignPose2(; mkd1::ManifoldKernelDensity, mkd2::ManifoldKernelDensity, moreoptions...)

Specialization of ScatterAlign for Pose2.

See also: ScatterAlignPose3

ScatterAlignPose3(; cloud1=mkd1::ManifoldKernelDensity, 
                    cloud2=mkd2::ManifoldKernelDensity, 
                    moreoptions...)

Specialization of ScatterAlign for Pose3.

See also: ScatterAlignPose2

alignICP_Simple(
    X_fix,
    X_mov;
    correspondences,
    neighbors,
    min_planarity,
    max_overlap_distance,
    min_change,
    max_iterations,
    verbose,
    H
)

Align two point clouds using ICP (with normals).

Example:

using Downloads, DelimitedFiles
using Colors, Caesar

# get some test data (~50mb download)
lidar1_url = "https://github.com/JuliaRobotics/CaesarTestData.jl/raw/main/data/lidar/simpleICP/terrestrial_lidar1.xyz"
lidar2_url = "https://github.com/JuliaRobotics/CaesarTestData.jl/raw/main/data/lidar/simpleICP/terrestrial_lidar2.xyz"
io1 = PipeBuffer()
io2 = PipeBuffer()
Downloads.download(lidar1_url, io1)
Downloads.download(lidar2_url, io2)

X_fix = readdlm(io1)
X_mov = readdlm(io2)

H, HX_mov, stat = Caesar._PCL.alignICP_Simple(X_fix, X_mov; verbose=true)

Notes

• Mostly consolidated with Caesar._PCL types.
• Internally uses Caesar._PCL._ICP_PointCloud which was created to help facilite consolidation of code:
  • Modified from www.github.com/pglira/simpleICP (July 2022).
• See here for a brief example on Visualizing Point Clouds.

DevNotes

• TODO switch rigid transfrom to Caesar._PCL.apply along with performance considerations, instead of current transform!.

See also: PointCloud