easynav_navmap_localizer package from easynav_plugins repo

easynav_costmap_common easynav_simple_common easynav_mpc_controller easynav_mppi_controller easynav_regulated_pp_controller easynav_serest_controller easynav_simple_controller easynav_vff_controller easynav_costmap_localizer easynav_fusion_localizer easynav_mhamcl_localizer easynav_navmap_localizer easynav_simple_localizer easynav_bonxai_maps_manager easynav_costmap_maps_manager easynav_navmap_maps_manager easynav_octomap_maps_manager easynav_routes_maps_manager easynav_simple_maps_manager easynav_costmap_planner easynav_navmap_planner easynav_simple_planner easynav_diagnostic_recovery easynav_simple_recovery

ROS Distro
humble

API Docs
Browse Code

Package Summary

Version	0.5.1
License	Apache-2.0
Build type	AMENT_CMAKE
Use	RECOMMENDED

Repository Summary

Description
Checkout URI	https://github.com/EasyNavigation/easynav_plugins.git
VCS Type	git
VCS Version	humble
Last Updated	2026-10-23
Dev Status	DEVELOPED
Released	RELEASED
Contributing	Help Wanted (-) Good First Issues (-) Pull Requests to Review (-)

Package Description

Easy Navigation: nAVmAP Localizer package.

Maintainers

Francisco Martín Rico

Authors

No additional authors.

localizers/easynav_navmap_localizer/README.md

easynav_navmap_localizer

Description

Easy Navigation: nAVmAP Localizer package.

Authors and Maintainers

Authors: Intelligent Robotics Lab
Maintainers: Francisco Martín Rico fmrico@gmail.com

Supported ROS 2 Distributions

Distribution	Status
humble
jazzy
kilted
lyrical
rolling

Plugin (pluginlib)

Plugin Name: easynav_navmap_localizer/AMCLLocalizer
Type: easynav::navmap::AMCLLocalizer
Base Class: easynav::LocalizerMethodBase
Library: easynav_navmap_localizer
Description: A default “navmap” implementation for the AMCL localizer.

Parameters

All parameters are declared under the plugin name namespace, i.e., /<node_fqn>/easynav_navmap_localizer/AMCLLocalizer/....

Name	Type	Default	Description
`<plugin>.num_particles`	`int`	`100`	Number of AMCL particles.
`<plugin>.initial_pose.x`	`double`	`0.0`	Initial X position (m).
`<plugin>.initial_pose.y`	`double`	`0.0`	Initial Y position (m).
`<plugin>.initial_pose.yaw`	`double`	`0.0`	Initial yaw (rad).
`<plugin>.initial_pose.std_dev_xy`	`double`	`0.5`	Std dev used to sample initial X/Y.
`<plugin>.initial_pose.std_dev_yaw`	`double`	`0.5`	Std dev used to sample initial yaw.
`<plugin>.reseed_freq`	`double`	`1.0`	Reseeding frequency (Hz).
`<plugin>.noise_translation`	`double`	`0.01`	Translational noise factor.
`<plugin>.noise_rotation`	`double`	`0.01`	Rotational noise factor.
`<plugin>.noise_translation_to_rotation`	`double`	`0.01`	Translation-to-rotation noise coupling.
`<plugin>.min_noise_xy`	`double`	`0.05`	Minimum XY noise (m).
`<plugin>.min_noise_yaw`	`double`	`0.05`	Minimum yaw noise (rad).
`<plugin>.compute_odom_from_tf`	`bool`	`false`	If true, read odometry from TF (odom->base_footprint) instead of /odom topic.
`<plugin>.inflation_stddev`	`double`	`0.05`	Std dev used in occupancy inflation (m).
`<plugin>.inflation_prob_min`	`double`	`0.01`	Minimum occupancy probability after inflation.
`<plugin>.correct_max_points`	`int`	`1500`	Maximum points of each sensor’s cloud scored per correction (after downsampling and the height band), evenly spread over it.
`<plugin>.weights_tau`	`double`	`0.7`	Exponential decay factor for particle weight smoothing.
`<plugin>.top_keep_fraction`	`double`	`0.2`	Fraction of top particles to keep before reseeding.
`<plugin>.downsampled_cloud_size`	`double`	`0.05`	Voxel size for point cloud downsampling (m). Of the order of the Bonxai map’s voxels: much larger leaves too few points to tell poses apart.
`<plugin>.min_point_height`	`double`	`-inf`	Only points at least this high over the robot’s footprint are scored (m). For 3D sensors: the floor fits any particle standing on it.
`<plugin>.max_point_height`	`double`	`inf`	Only points at most this high over the robot’s footprint are scored (m), e.g. below a ceiling.
`<plugin>.ray_occlusion`	`bool`	`false`	A point whose ray from the sensor crosses an occupied voxel scores nothing. With voxels a few decimetres wide, rays along a floor or a ceiling look occluded, and the true pose scores worse than wrong ones.
`<plugin>.reseed_min_d`	`double`	`0.1`	Reseed only after the odometry has moved this much since the last reseed (m), or turned `reseed_min_a`: standing still, reseeding makes the particles wander.
`<plugin>.reseed_min_a`	`double`	`0.1`	See `reseed_min_d` (rad).
`<plugin>.odom_slip_fraction`	`double`	`0.05`	Each particle moves with its own scale of the odometry’s translation. This fraction of the reseeded ones take any scale in [0, 1]: when the wheels slip, those that hardly move fit the sensors and survive. Rough terrain: higher (0.2).
`<plugin>.odom_scale_noise`	`double`	`0.02`	Noise of the scale a reseeded particle takes from its reference: the particles learn the odometry’s scale.
`<plugin>.recovery_alpha_slow`	`double`	`0.001`	Rate of the slow average of the scan likelihood (of the best particles), per correction.
`<plugin>.recovery_alpha_fast`	`double`	`0.1`	Rate of the fast average. When it falls below the slow one, each reseed spreads the worst particles widely around the estimate, a fraction `1 - fast/slow` of those it replaces. `0` for both turns recovery off; it needs `0 <= slow < fast <= 1`.
`<plugin>.recovery_spread_xy`	`double`	`1.0`	Standard deviation of the spread particles’ position around the estimate (m).
`<plugin>.recovery_spread_yaw`	`double`	`0.5`	Standard deviation of their heading around the best particle’s (rad).

Interfaces (Topics and Services)

Subscriptions and Publications

Direction	Topic	Type	Purpose	QoS
Subscription	`/odom`	`nav_msgs/msg/Odometry`	Used only when `<plugin>.compute_odom_from_tf = false`	SensorDataQoS (reliable)
Subscription	`initialpose`	`geometry_msgs/msg/PoseWithCovarianceStamped`	Initialize particles pose to the received pose, using covariance	depth=10
Publisher	`<node_fqn>/<plugin>/particles`	`geometry_msgs/msg/PoseArray`	Publishes the current particle set	depth=10
Publisher	`<node_fqn>/<plugin>/pose`	`geometry_msgs/msg/PoseWithCovarianceStamped`	Publishes the estimated pose with covariance	depth=10

Services

This package does not create service servers or clients.

NavState Keys

Key	Type	Access	Notes
`imu`	`IMUPerceptions`	Read	When available, the particles take its roll and pitch, and turn by the change of its yaw (not by the odometry’s); its absolute yaw is never used (without a magnetometer it is arbitrary and drifts). Without it, the orientation follows the NavMap surface and the odometry’s turns.
`points`	`PointPerceptions`	Read	Point clouds used in the correction step: only the valid ones (newer than the sensors’ `forget_time`), each where the sensor was when it took it (the odometry since then, through `odom`).
`map.navmap`	`::navmap::NavMap`	Read	Triangle mesh map providing surface geometry and elevation.
`map.bonxai`	`Bonxai::ProbabilisticMap`	Read	Probabilistic occupancy map used for scoring.
`map.bonxai.inflated`	`Bonxai::ProbabilisticMap`	Write/Read	Inflated map cached on first access for faster scoring.
`robot_pose`	`nav_msgs::msg::Odometry`	Write	Estimated robot pose stored at the end of update/predict, with its roll and pitch (the IMU’s, or the surface’s).

TF Frames

Role	Transform	Notes
Publishes	`map -> odom`	Broadcasts the global transform that aligns the odometry frame with the map frame. It carries the robot’s roll and pitch (the IMU’s, or the surface’s) when the odometry is planar: on a slope, the points the sensors see are placed where they are.
Requires	`odom -> base_footprint`	Required only when `<plugin>.compute_odom_from_tf = true` (read from TF).
Requires	`base_footprint -> <sensor_frame>`	Required to transform perception point clouds into the robot base frame during correction, at the cloud’s time through `odom` (falls back to the latest one).

File truncated at 100 lines see the full file

CHANGELOG

Changelog for package easynav_navmap_localizer

0.5.1 (2026-10-21)

Robust in 3D: estimate with roll and pitch, height band of scored points, motion-compensated clouds, per-particle odometry scale for wheel slip
Recovery from a wrong convergence (augmented MCL)
Lock-free split between the RT and non-RT threads
Per-particle heading noise; the IMU gives turns, not its absolute yaw
Do not force the Debug log level
Enable cpplint and copyright linters
READMEs: supported on Humble, Jazzy, Kilted, Lyrical and Rolling
Contributors: Francisco Martín Rico

0.5.0 (2026-10-08)

Runtime reconfiguration
Fixed the particle distribution
Builds on Humble, Jazzy, Kilted, Lyrical and Rolling
Contributors: Francisco Martín Rico

0.4.2 (2026-07-26)

Complete deps
Check if std is non-positive before creating std::normal_distribution
Update README.md for all localizers
Reset pose from RViz2 for every localizer
Adaptations to #94
Update plugins to new sensors API
GPLv3 -> Apache 2.0
Improve navstate print including the time
Add a base_footprint frame in TFInfo
Remove C++20/C++23 features and update to new MethodBase interface
Correct publication stamp to input stamp
Merge branch 'set_robot_frame' into frames-fix-pr-40
TFInfo in RTTFBuffer
Refactor to use TFInfo
Optimize execution
Optimize random genearator
Optimize execution
Update sheets
Contributors: Francisco Martín Rico, Francisco Miguel Moreno, Juan S. Cely, Juan S. Cely G., Miguel, migueldm

0.0.2 (2025-10-12)

Reorganization initial
Contributors: Francisco Martín Rico

Package Dependencies

Deps	Name
	ament_cmake
	ament_cmake_gtest
	ament_lint_auto
	ament_lint_common
	rclcpp_lifecycle
	std_srvs
	easynav_bonxai_maps_manager
	easynav_common
	easynav_core
	easynav_localizer
	easynav_sensors
	geometry_msgs
	nav_msgs
	navmap_core
	navmap_ros
	navmap_ros_interfaces
	pluginlib
	rclcpp
	tf2
	tf2_geometry_msgs
	tf2_ros

System Dependencies

Name
eigen
libpcl-all-dev

Dependant Packages

No known dependants.

Launch files

No launch files found

Messages

No message files found.

Services

No service files found

Plugins

No plugins found.

Recent questions tagged `easynav_navmap_localizer` at Robotics Stack Exchange

API Docs Browse Code