cuboidModel

Parametric cuboid model

Since R2020b

Description

The cuboidModel object stores the parameters of a parametric cuboid model. After you create a cuboidModel object, you can extract cuboid corner points, and points within the cuboid using the object functions. Cuboid models are used to store the output of pcfitcuboid function. It is a shape fitting function which fits a cuboid over a point cloud.

Creation

There are two ways to create a cuboidModel object.

• Create a cuboid model by specifying the cuboid parameters in the cuboidModel function.

• Fit a cuboid model over a point cloud using the pcfitcuboid function.

Description

model = cuboidModel(params) constructs a parametric cuboid model from the 1-by-9 input vector, params.

model = pcfitcuboid(ptCloudIn) fits a cuboid over the input point cloud data. The pcfitcuboid function stores the properties of the cuboid in a parametric cuboid model object, model.

model = pcfitcuboid(ptCloudIn,indices) fits a cuboid over a selected set of points, indices, in the input point cloud.

Properties

Parameters — Cuboid model parameters

nine-element row vector

This property is read-only.

Cuboid model parameters, stored as a nine-element row vector of the form [x_ctr y_ctr z_ctr x_len y_len z_len x_rot y_rot z_rot].

• x_ctr, y_ctr, and z_ctr specify the center of the cuboid.

• x_len, y_len, and z_len specify the length of the cuboid along the x-, y-, and z-axis, respectively, before rotation has been applied.

• x_rot, y_rot, and z_rot specify the rotation angles in degrees for the cuboid along the x-, y-, and z-axis, respectively. These angles are clockwise-positive when looking in the forward direction of their corresponding axes.

The figure shows how these values determine the position of a cuboid.

These parameters are specified by the params input argument.

Data Types: single | double

Center — Center of cuboid

three-element row vector

This property is read-only.

Center of the cuboid, stored as a three-element row vector of the form [x_ctr y_ctr z_ctr]. The vector contains the 3-D coordinates of the cuboid center in the x-, y-, and z-axis, respectively.

This property is derived from the Parameters property.

Data Types: single | double

Dimensions — Dimensions of cuboid

three-element row vector

This property is read-only.

Dimensions of the cuboid, stored as a three-element row vector of the form [x_len y_len z_len]. The vector contains the length of the cuboid along the x-, y-, and z-axis, respectively.

This property is derived from the Parameters property.

Data Types: single | double

Orientation — Orientation of cuboid

three-element row vector

This property is read-only.

Orientation of the cuboid, stored as a three-element row vector of the form, [x_rot y_rot z_rot], in degrees. The vector contains the rotation of the cuboid along the x-, y-, and z-axis, respectively.

This property is derived from the Parameters property.

If the orientation is in quaternion, convert the quaternion to Euler angles in degrees to create a cuboid model.

• To convert a quaternion to Euler angles in radians, use the quat2eul (Robotics System Toolbox) function. Set the sequence (Robotics System Toolbox) argument of the quat2eul (Robotics System Toolbox) function to "XYZ".

• To convert angle units from radians to degrees, use the rad2deg function.

Data Types: single | double

Object Functions

getCornerPoints	Get corner points of cuboid model
findPointsInModel	Find points enclosed by cuboid model
plot	Plot cuboid model

Examples

Detect Cuboid in Point Cloud

Detect a cuboid in a point cloud using pcfitcuboid function. The function stores the cuboid parameters as a cuboidModel object.

Read point cloud data into the workspace.

ptCloud = pcread('highwayScene.pcd');

Search the point cloud within a specified region of interest (ROI). Create a point cloud of only the detected points.

roi = [-30 30 -20 30 -8 13];
in = findPointsInROI(ptCloud,roi);
ptCloudIn = select(ptCloud,in);

Plot the point cloud of detected points.

   figure
   pcshow(ptCloudIn.Location)
   xlabel('X(m)')
   ylabel('Y(m)')
   zlabel('Z(m)')
   title('Detected Points in ROI')

Find the indices of the points in a specified ROI within the point cloud.

roi = [9.6 13.8 7.9 9.3 -2.5 3];
sampleIndices = findPointsInROI(ptCloudIn,roi);

Fit a cuboid to the selected set of points in the point cloud.

 model = pcfitcuboid(ptCloudIn,sampleIndices);
  figure
  pcshow(ptCloudIn.Location)
  xlabel('X(m)')
  ylabel('Y(m)')
  zlabel('Z(m)')
  title('Detect a Cuboid in a Point Cloud')

Plot the cuboid box in the point cloud.

hold on
plot(model)

Display the internal properties of the cuboidModel object.

model

model = 
  cuboidModel with properties:

     Parameters: [11.4873 8.5997 -1.6138 3.6713 1.3220 1.7576 0 0 0.9999]
         Center: [11.4873 8.5997 -1.6138]
     Dimensions: [3.6713 1.3220 1.7576]
    Orientation: [0 0 0.9999]

Fit Cuboid Over Point Cloud Data

Fit cuboid bounding boxes around clusters in a point cloud.

Load the point cloud data into the workspace.

data = load('drivingLidarPoints.mat');

Define and crop a region of interest (ROI) from the point cloud. Visualize the selected ROI of the point cloud.

roi = [-40 40 -6 9 -2 1];
in = findPointsInROI(data.ptCloud,roi);
ptCloudIn = select(data.ptCloud,in);
hcluster = figure;
panel = uipanel('Parent',hcluster,'BackgroundColor',[0 0 0]);
ax = axes('Parent',panel,'Color',[0 0 0]); 
pcshow(ptCloudIn,'MarkerSize',30,'Parent',ax)
title('Input Point Cloud')

Segment the ground plane. Visualize the segmented ground plane.

maxDistance = 0.3;
referenceVector = [0 0 1];
[~,inliers,outliers] = pcfitplane(ptCloudIn,maxDistance,referenceVector);
ptCloudWithoutGround = select(ptCloudIn,outliers,'OutputSize','full');
hSegment = figure;
panel = uipanel('Parent',hSegment,'BackgroundColor',[0 0 0]);
ax = axes('Parent',panel,'Color',[0 0 0]); 
pcshowpair(ptCloudIn,ptCloudWithoutGround,'Parent',ax)
legend('Ground Region','Non-Ground Region','TextColor', [1 1 1])
title('Segmented Ground Plane')

Segment the non-ground region of the point cloud into clusters. Visualize the segmented point cloud.

distThreshold = 1;
[labels,numClusters] = pcsegdist(ptCloudWithoutGround,distThreshold);
labelColorIndex = labels;
hCuboid = figure;
panel = uipanel('Parent',hCuboid,'BackgroundColor',[0 0 0]);
ax = axes('Parent',panel,'Color',[0 0 0]); 
pcshow(ptCloudIn.Location,labelColorIndex,'Parent',ax)
title('Fitting Bounding Boxes')
hold on

Fit bounding box on each cluster, visualized as orange highlights.

for i = 1:numClusters
    idx = find(labels == i);
    model = pcfitcuboid(ptCloudWithoutGround,idx);
    plot(model)
end

Extended Capabilities

C/C++ Code Generation
Generate C and C++ code using MATLAB® Coder™.

Version History

Introduced in R2020b