Collaboration diagram for apollo::perception::lidar::PreprocessPointsCuda:

Public Member Functions
	PreprocessPointsCuda (const int num_threads, const int max_num_pillars, const int max_points_per_pillar, const int num_point_feature, const int num_inds_for_scan, const int grid_x_size, const int grid_y_size, const int grid_z_size, const float pillar_x_size, const float pillar_y_size, const float pillar_z_size, const float min_x_range, const float min_y_range, const float min_z_range)
	Constructor. More...

	~PreprocessPointsCuda ()

void	DoPreprocessPointsCuda (const float dev_points, const int in_num_points, int dev_x_coors, int dev_y_coors, float dev_num_points_per_pillar, float dev_pillar_point_feature, float dev_pillar_coors, int dev_sparse_pillar_map, int host_pillar_count)
	CUDA preprocessing for input point cloud. More...

Constructor & Destructor Documentation

◆ PreprocessPointsCuda()

apollo::perception::lidar::PreprocessPointsCuda::PreprocessPointsCuda	(	const int	num_threads,
		const int	max_num_pillars,
		const int	max_points_per_pillar,
		const int	num_point_feature,
		const int	num_inds_for_scan,
		const int	grid_x_size,
		const int	grid_y_size,
		const int	grid_z_size,
		const float	pillar_x_size,
		const float	pillar_y_size,
		const float	pillar_z_size,
		const float	min_x_range,
		const float	min_y_range,
		const float	min_z_range
	)

Constructor.

Parameters

[in]	num_threads	Number of threads when launching cuda kernel
[in]	max_num_pillars	Maximum number of pillars
[in]	max_points_per_pillar	Maximum number of points per pillar
[in]	num_point_feature	Number of features in a point
[in]	num_inds_for_scan	Number of indexes for scan(cumsum)
[in]	grid_x_size	Number of pillars in x-coordinate
[in]	grid_y_size	Number of pillars in y-coordinate
[in]	grid_z_size	Number of pillars in z-coordinate
[in]	pillar_x_size	Size of x-dimension for a pillar
[in]	pillar_y_size	Size of y-dimension for a pillar
[in]	pillar_z_size	Size of z-dimension for a pillar
[in]	min_x_range	Minimum x value for point cloud
[in]	min_y_range	Minimum y value for point cloud
[in]	min_z_range	Minimum z value for point cloud

Captital variables never change after the compile

apollo::perception::lidar::PreprocessPointsCuda::~PreprocessPointsCuda ( )

void apollo::perception::lidar::PreprocessPointsCuda::DoPreprocessPointsCuda	(	const float *	dev_points,
		const int	in_num_points,
		int *	dev_x_coors,
		int *	dev_y_coors,
		float *	dev_num_points_per_pillar,
		float *	dev_pillar_point_feature,
		float *	dev_pillar_coors,
		int *	dev_sparse_pillar_map,
		int *	host_pillar_count
	)

CUDA preprocessing for input point cloud.

Parameters

[in]	dev_points	Point cloud array
[in]	in_num_points	The number of points
[in]	dev_x_coors	X-coordinate indexes for corresponding pillars
[in]	dev_y_coors	Y-coordinate indexes for corresponding pillars
[in]	dev_num_points_per_pillar	Number of points in corresponding pillars
[in]	pillar_point_feature	Values of point feature in each pillar
[in]	pillar_coors	Array for coors of pillars
[in]	dev_sparse_pillar_map	Grid map representation for pillar-occupancy
[in]	host_pillar_count	The number of valid pillars for an input point cloud

Convert point cloud to pillar representation

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

modules/perception/lidar/lib/detector/point_pillars_detection/preprocess_points_cuda.h