Skip to content

Latest commit

 

History

83 Commits

Folders and files

NameName
Last commit message
Last commit date
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Passive Stereo ROS2

This repository provides a simple ROS 2 pipeline for passive stereo depth estimation using Retinify and conversion to 3D colored point clouds.


Features

  • 🔹 Disparity Estimation A node that runs the Retinify disparity network on rectified stereo pairs and publishes the disparity image.

  • 🔹 Disparity → PointCloud2 A node that converts the disparity map into a sensor_msgs/PointCloud2, using camera calibration parameters. The left rectified image is used as the RGB texture of the 3D point cloud.

  • 🔹 ROS 2 native (tested on Humble + JetPack 6.2 / x86_64).


Dependencies


Installation

Clone into your workspace:

cd ~/ros2_ws/src
git clone https://github.com/Projeto-Voris/passive_stereo.git
cd ..
colcon build
source install/setup.bash

Nodes

retinify_disp

Runs Retinify disparity estimation.

Subscribed topics:

  • /stereo/left/image_raw (sensor_msgs/Image)
  • /stereo/right/image_raw (sensor_msgs/Image)
  • /stereo/left/camera_info (sensor_msgs/CameraInfo)
  • /stereo/right/camera_info (sensor_msgs/CameraInfo)

Published topics:

  • /stereo/disparity/image (stereo_msgs/DisparityImage)
  • /stereo/disparity/debug/image (sensor_msgs/Image)
  • /stereo/left/rect_image (sensor_msgs/Image)
  • /stereo/right/rect_image (sensor_msgs/Image)

Parameters

Parameter Description Value
publish_rectified Publish rectified images True
debug_image Publish debug disparity image False

Triangulation

Converts disparity to colored point cloud from left rectified image.

Subscribed topics:

  • /stereo/disparity (stereo_msgs/DisparityImage)
  • /stereo/left/image_rect (sensor_msgs/Image)
  • /stereo/left/camera_info (sensor_msgs/CameraInfo)

Published topics:

  • /stereo/points2 (sensor_msgs/PointCloud2)

Parameters

Parameter Description Value
frame_id Pointcloud frame ID left_camera_link
sampling_factor Publish debug disparity image 0.8
crop_factor Maximum occupancy threshold 0.9

Usage

  1. Launch your stereo camera drivers (e.g. FLIR, RealSense, ZED) that publish raw images and camera info.

  2. Run:

    ros2 launch passive_stereo passive_stereo.launch.py

License

MIT License. See LICENSE for details.

About

ROS2 Humble package to calculate a image pair disparity

Resources

Stars

0 stars

Watchers

0 watching

Forks

Releases

Packages

Contributors

Languages