autoware_downsample_filters#

Overview#

The autoware_downsample_filters is a package that that reduces the number of points of the pointcloud.

Design#

The autoware_downsample_filters is implemented for reducing the computational load and improving the performance of the system.

Inner-workings / Algorithms#

Random Downsample Filter#

pcl::RandomSample is used, which points are sampled with uniform probability.

Voxel Grid Downsample Filter#

pcl::VoxelGrid is used, which points in each voxel are approximated with their centroid.

Pickup Based Voxel Grid Downsample Filter#

This algorithm samples a single actual point existing within the voxel, not the centroid. The computation cost is low compared to Centroid Based Voxel Grid Filter.

Inputs / Outputs#

Input#

Name	Type	Description
`input`	`sensor_msgs::msg::PointCloud2`	reference points

Output#

Name	Type	Description
`output`	`sensor_msgs::msg::PointCloud2`	downsampled points

Parameters#

Launch file Parameters#

Name	Type	Default Value	Description
`input_frame`	string	" "	the frame id in which filtering is performed
`output_frame`	string	" "	output frame id of the filtered points
`max_queue_size`	size_t	5	max buffer size of input/output topics

Node Parameters#

random_downsample_filter_node#

Name	Type	Default Value	Description
`sample_num`	size_t	1500	random sample number

voxel_grid_downsample_filter_node#

Name	Type	Default Value	Description
`voxel_size_x`	double	0.3	x value of the voxel
`voxel_size_y`	double	0.3	y value of the voxel
`voxel_size_z`	double	0.1	z value of the voxel

Usage#

1.publish static tf from input pointcloud to target frame that is used for filtering, e.g#

ros2 run tf2_ros static_transform_publisher 2.0 3.2 1.3 0 0 0 1  velodyne_top_base_link  base_link

2.launch one of the downsample node#

ros2 launch autoware_downsample_filters random_downsample_filter_node.launch.xml