No version for distro humble showing github. Known supported distros are highlighted in the buttons above.

semantic_segmentation_node package from navigation2_tutorials repo

nav2_costmap_filters_demo nav2_gps_waypoint_follower_demo nav2_gradient_costmap_plugin nav2_pure_pursuit_controller semantic_segmentation_node semantic_segmentation_sim nav2_sms_behavior nav2_straightline_planner sam_bot_description

ROS Distro
github

API Docs
Browse Code

Package Summary

Version	0.0.1
License	BSD-3-Clause
Build type	AMENT_PYTHON
Use	RECOMMENDED

Repository Summary

Description	Tutorial code referenced in https://docs.nav2.org/
Checkout URI	https://github.com/ros-navigation/navigation2_tutorials.git
VCS Type	git
VCS Version	rolling
Last Updated	2026-02-20
Dev Status	UNKNOWN
Released	UNRELEASED
Contributing	Help Wanted (-) Good First Issues (-) Pull Requests to Review (-)

Package Description

ROS2 node for semantic segmentation inference

Maintainers

Pedro Gonzalez

Authors

No additional authors.

nav2_semantic_segmentation_demo/semantic_segmentation_node/README.md

Semantic Segmentation Node

ROS2 node for real-time semantic segmentation inference using ONNX Runtime.

Overview

This node performs semantic segmentation on camera images and publishes segmentation masks, confidence maps, and colored overlays. It uses ONNX Runtime for efficient inference without requiring PyTorch or super-gradients at runtime.

Topics

Subscribed:

/rgbd_camera/image (sensor_msgs/Image) - Input RGB camera images

Published:

/segmentation/mask (sensor_msgs/Image) - Segmentation mask with class IDs (mono8)
/segmentation/confidence (sensor_msgs/Image) - Per-pixel confidence (mono8, 0-255)
/segmentation/overlay (sensor_msgs/Image) - Colored overlay visualization (bgr8)
/segmentation/label_info (vision_msgs/LabelInfo) - Class mappings (latched)

Model

The ONNX model (models/model.onnx) can be generated using the Simple Segmentation Toolkit.

Training Your Own Model

Capture training images from a real robot or from Gazebo, with varying lighting and environmental conditions
Use the Simple Segmentation Toolkit to label and train a model
Convert the trained model to ONNX format: python3 convert_to_onnx.py
Copy model.onnx to this package’s models/ directory

The ontology configuration (config/ontology.yaml) must match the classes used during training.

Usage

ros2 run semantic_segmentation_node segmentation_node

All dependencies are included in the devcontainer.

CHANGELOG

No CHANGELOG found.

Package Dependencies

Deps	Name
	rclpy
	sensor_msgs
	cv_bridge
	std_msgs
	vision_msgs

System Dependencies

No direct system dependencies.

Dependant Packages

Name	Deps
semantic_segmentation_sim

Launch files

No launch files found

Messages

No message files found.

Services

No service files found

Plugins

No plugins found.

Recent questions tagged `semantic_segmentation_node` at Robotics Stack Exchange

API Docs Browse Code

No version for distro jazzy showing github. Known supported distros are highlighted in the buttons above.