pcregisterfgr

Register two point clouds using FGR algorithm

Since R2022b

Syntax

tform = pcregisterfgr(moving,fixed,gridSize)

[tform,rmse] = pcregisterfgr(moving,fixed,gridSize)

[___] = pcregisterfgr(___,MaxIterations=numIterations)

Description

tform = pcregisterfgr(moving,fixed,gridSize) returns a rigid transformation that registers a moving point cloud to a fixed point cloud.

The function registers points using a fast global registration (FGR) algorithm based on FPFH features.

[tform,rmse] = pcregisterfgr(moving,fixed,gridSize) additionally returns the root mean square error of the Euclidean distance between the inlier aligned points.

[___] = pcregisterfgr(___,MaxIterations=numIterations) specifies the maximum number of iterations for the FGR algorithm, in addition to any combination of arguments from previous syntaxes.

Examples

Align Two Point Clouds Using FGR Algorithm

Load point cloud data into the workspace.

ld = load("livingRoom.mat");
fixed = ld.livingRoomData{1};
moving = ld.livingRoomData{2};

Visualize point cloud data before registration.

figure
pcshowpair(moving,fixed,VerticalAxis="Y",VerticalAxisDir="Down") 

To improve the efficiency and the accuracy of the FGR registration algorithm, downsample the point clouds.

fixedDownsampled = pcdownsample(fixed,gridNearest=0.05);
movingDownsampled = pcdownsample(moving,gridNearest=0.05);

Register the point clouds using the FGR algorithm.

gridSize = 0.1;
tform = pcregisterfgr(movingDownsampled,fixedDownsampled,gridSize,MaxIterations=100);

Visualize the alignment.

movingRegistered = pctransform(moving,tform);
pcshowpair(movingRegistered,fixed,VerticalAxis="Y",VerticalAxisDir="Down");

Merge the point clouds and visualize the merged result.

mergeSize = 0.01;
merged = pcmerge(movingRegistered,fixed,mergeSize);
figure(Name = "merged point cloud after Fast Global registration");
pcshow(merged,VerticalAxis="Y",VerticalAxisDir ="Down");

Input Arguments

moving — Moving point cloud

pointCloud object

Moving point cloud, specified as a pointCloud object.

fixed — Fixed point cloud

pointCloud object

Fixed point cloud, specified as a pointCloud object.

gridSize — Grid size to search for correspondence between point clouds

positive scalar

Grid size to search for correspondence between the point clouds, specified as a positive scalar.

Data Types: single | double | int8 | int16 | int32 | int64 | uint8 | uint16 | uint32 | uint64

numIterations — Maximum number of iterations

64 (default) | positive integer

Maximum number of iterations before the FGR algorithm stops, specified as a positive integer.

Data Types: single | double | int8 | int16 | int32 | int64 | uint8 | uint16 | uint32 | uint64

Output Arguments

tform — Rigid transformation

rigidtform3d object

Rigid transformation, returned as a rigidtform3d object. The rigid transformation registers a moving point cloud to a fixed point cloud. The rigidtform3d object describes the rigid 3-D transform. The FGR algorithm estimates the rigid transformation between the moving and fixed point clouds.

rmse — Root mean square error

positive scalar

Root mean square error, returned as a positive scalar that represents the Euclidean distance between the inlier aligned points. A lower error values indicates a better registration.

Tips

• To improve the accuracy and the efficiency of registration, downsample the point clouds using the pcdownsample function before using the pcregisterfgr function.

• For ground vehicle point clouds, you can improve performance and accuracy by removing the ground using pcfitplane or segmentGroundFromLidarData before registration. For details on how to do this, see the helperProcessPointCloud function in the Build a Map from Lidar Data (Automated Driving Toolbox) example.

• To merge more than two point clouds, you can use the pccat function instead of the pcmerge function.

Algorithms

This figure shows the workflow of point cloud registration using the FGR algorithm.

Version History

Introduced in R2022b