autoware::perception::tracking::GreedyRoiAssociator Class Reference

Class to associate the detections and tracks in euclidean space to ROIs in image space on a first-come-first-serve manner. More...

#include <greedy_roi_associator.hpp>

Public Types
using	float32_t = autoware::common::types::float32_t

Public Member Functions
	GreedyRoiAssociator (const GreedyRoiAssociatorConfig &config, const tf2::BufferCore &tf_buffer)
	Constructor. More...
AssociatorResult	assign (const autoware_auto_perception_msgs::msg::ClassifiedRoiArray &rois, const TrackedObjects &tracks) const
	Assign the tracks to the ROIs. The assignment is done by first projecting the tracks, Then assigning each track to a detection according to the IoU metric in a greedy fashion. More...
AssociatorResult	assign (const autoware_auto_perception_msgs::msg::ClassifiedRoiArray &rois, const autoware_auto_perception_msgs::msg::DetectedObjects &objects) const
	Assign the objects to the ROIs. The assignment is done by first projecting the detections, then assigning each detection to a ROI according to the IoU metric in a greedy fashion. More...

Detailed Description

Class to associate the detections and tracks in euclidean space to ROIs in image space on a first-come-first-serve manner.

Member Typedef Documentation

float32_t

using autoware::perception::tracking::GreedyRoiAssociator::float32_t = autoware::common::types::float32_t

Constructor & Destructor Documentation

GreedyRoiAssociator()

autoware::perception::tracking::GreedyRoiAssociator::GreedyRoiAssociator ( const GreedyRoiAssociatorConfig &  config, const tf2::BufferCore &  tf_buffer  )

explicit

Constructor.

Parameters

config	Configuration for projection and association
tf_buffer	TF2 buffer

Member Function Documentation

assign() [1/2]

AssociatorResult autoware::perception::tracking::GreedyRoiAssociator::assign	(	const autoware_auto_perception_msgs::msg::ClassifiedRoiArray &	rois,
		const autoware_auto_perception_msgs::msg::DetectedObjects &	objects
	)		const

Assign the objects to the ROIs. The assignment is done by first projecting the detections, then assigning each detection to a ROI according to the IoU metric in a greedy fashion.

Parameters

rois	Regions of Interest in camera frame from vision subsystem
objects	DetectedObjects from lidar or radar

Returns

The association between the objects and the rois. In this case, "tracks" in the return struct refers to the 3D objects and "detections" refer to the ROIs

assign() [2/2]

AssociatorResult autoware::perception::tracking::GreedyRoiAssociator::assign	(	const autoware_auto_perception_msgs::msg::ClassifiedRoiArray &	rois,
		const TrackedObjects &	tracks
	)		const

Assign the tracks to the ROIs. The assignment is done by first projecting the tracks, Then assigning each track to a detection according to the IoU metric in a greedy fashion.

Parameters

rois	ROI detections
tracks	Tracks

Returns

The association between the tracks and the rois

---

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

• greedy_roi_associator.hpp
• greedy_roi_associator.cpp