Event-based ORB-SLAM3 ROS Node

Overview

This project is a ROS1-based implementation of an event-based SLAM system that integrates rpg_dvs_ros and EORB_SLAM to process event camera data. The node operates on Ubuntu 20.04 with ROS Noetic and is designed to handle event-based datasets, particularly ETHZ Public Event Dataset. It utilizes ORB-SLAM3 for state estimation and event tracking.

References

This project integrates components from:

• EORB_SLAM: An event-based extension of ORB-SLAM3, developed by Masoud Dayani Najafabadi.

Features

• Event-based SLAM with ORB-SLAM3 integration.
• Time-synchronized event processing with batch thresholding.
• Event timestamp normalization and sorting for improved accuracy.
• Adaptive event batching to optimize SLAM input.
• Event-based mapping and tracking for 6-DOF camera pose estimation.
• Multi-threaded implementation to ensure real-time performance.

Dependencies

Make sure you have the following dependencies installed:

• ROS Noetic (Ubuntu 20.04)
• rpg_dvs_ros (for event camera integration)
• EORB_SLAM (event-based ORB-SLAM3 extension)
• OpenCV 3.4.1
• PCL (Point Cloud Library)
• cv_bridge, image_transport, message_filters (ROS message handling)

Installation

1. Clone the required repositories:

   cd ~/catkin_ws/src
   git clone https://github.com/uzh-rpg/rpg_dvs_ros.git
   git clone https://github.com/m-dayani/EORB_SLAM.git

2. Install dependencies:

   sudo apt update && sudo apt install ros-noetic-cv-bridge ros-noetic-image-transport ros-noetic-message-filters libpcl-dev

3. Build the workspace:

   cd ~/catkin_ws
   catkin_make
   source devel/setup.bash

Usage

Run the EORB-SLAM node with an event dataset:

roslaunch eorb_slam3 eorbslam.launch settings:=/path/to/settings.yaml

Event Processing Pipeline

1. Event Data Handling:

   • Events are collected in vEvBuff.
   • If event timestamps are too close, they are accumulated before being sent to SLAM.
   • The batch threshold ensures event accumulation before processing.

2. SLAM Initialization:

   • Loads ORB Vocabulary from EORB_SLAM.
   • Reads dataset configurations from EvEthzLoader.

3. Event SLAM Execution:

   • ORB-SLAM3::System processes the incoming event stream.
   • Tracks event features, builds a map, and estimates the camera pose.
   • Outputs tracking state and map point statistics.

4. Data Logging:

   • Saves trajectory and SLAM-generated map:

   /home/user/trajectory.txt
   /home/user/orb_map.txt

Configuration File: EvETHZ.yaml

Key parameters configured in EvETHZ.yaml:

• Dataset Path:

  DS.name: "ev_ethz"
  Path.DS.root: "/ds_path/event/ethz"

• Camera Calibration:

  Camera.fx: 199.092
  Camera.fy: 198.828
  Camera.cx: 132.192
  Camera.cy: 110.712

• Event Windowing Parameters:

  Event.data.l1ChunkSize: 2000
  Event.data.minEvGenRate: 1.0
  Event.data.maxPixelDisp: 3.0

Acknowledgments

This project integrates components from:

• rpg_dvs_ros for event camera processing.
• EORB_SLAM for event-based SLAM implementation.

Troubleshooting Installation Issues for ORB-SLAM3 with ROS

Overview

This guide helps in solving common installation issues while setting up ORB-SLAM3 with ROS for event-based SLAM using Dynamic Vision Sensors (DVS). If you're encountering issues related to missing dependencies, incorrect configurations, or build problems, this guide will walk you through troubleshooting and resolving them.

Pre-requisites

Before starting, ensure you have the following installed:

• ROS Noetic (or compatible ROS version)
• ORB-SLAM3 installed on your system
• DVS/Camera Drivers (if using DAVIS or event cameras)
• OpenCV (version 4.x, compatible with ROS Noetic)
• Eigen, CMake, and other dependencies
• Catkin tools for building ROS packages

---

Troubleshooting Installation Issues for ORB-SLAM3 with ROS

Overview

This guide provides solutions to common installation issues encountered while setting up ORB-SLAM3 with ROS for event-based SLAM using Dynamic Vision Sensors (DVS). If you're experiencing problems related to missing dependencies, incorrect configurations, or build errors, follow this guide for troubleshooting.

Pre-requisites

Before starting, ensure that you have the following installed:

• ROS Noetic (or a compatible ROS version)
• ORB-SLAM3 installed in your system
• DVS/Camera Drivers (if using DAVIS or other event cameras)
• OpenCV (version 4.x, compatible with ROS Noetic)
• Eigen, CMake, and other dependencies
• Catkin tools for building ROS packages

Common Installation Issues & Solutions

Issue 1: Missing ROS Package orb_slam3

Error:

rosrun orb_slam3 orbslam_node
[ERROR] [ros]: Could not find package orb_slam3

Fix:

Ensure that ORB_SLAM3 is cloned into catkin_ws/src and properly built:

cd ~/catkin_ws/src
git clone https://github.com/UZH-RPG/ORB_SLAM3.git
cd ~/catkin_ws
catkin_make
source devel/setup.bash

Issue 2: Missing OpenCV Header Files or Linking Errors

Error:

fatal error: opencv2/core/core.hpp: No such file or directory

Fix:

Ensure OpenCV is installed and linked correctly:

sudo apt-get update
sudo apt-get install libopencv-dev

Issue 3: Missing catkin_make Dependencies

Error:

CMake Error at /opt/ros/noetic/share/catkin/cmake/catkinConfig.cmake:125 (find_package): Could not find a package configuration file provided by "ORB_SLAM3"

Fix:

Make sure ORB_SLAM3 is inside catkin_ws/src, then run:

source ~/catkin_ws/devel/setup.bash
catkin_make

Issue 4: CMake Errors with Eigen Library

Error:

CMake Error at CMakeLists.txt:13 (find_package): Could not find a package configuration file provided by "Eigen3"

Fix:

Install the missing Eigen3 library:

sudo apt-get install libeigen3-dev

If the error persists, manually link Eigen3 in your CMakeLists.txt.

Issue 5: Camera or IMU Driver Errors

Error:

[ERROR] Cannot load settings file.

Fix:

• Verify the event camera driver is correctly installed.
• Ensure the dataset path is correctly set in the YAML file.
• Check that the camera topics are correctly published (/dvs/image_raw, /dvs/events, /dvs/imu).

Issue 6: Missing ROS Dependencies (dvs_msgs, sensor_msgs, etc.)

Error:

[rosrun] Unable to locate package dvs_msgs

Fix:

Install missing ROS message packages:

sudo apt-get install ros-noetic-dvs-msgs
sudo apt-get install ros-noetic-sensor-msgs

Then, rebuild your workspace:

cd ~/catkin_ws
catkin_make
source devel/setup.bash

Issue 7: CMake Version Compatibility

Error:

CMake Error at CMakeLists.txt:16 (message): CMake version >= 3.10 is required.

Fix:

Ensure you have the correct version of CMake installed:

sudo apt-get install cmake

If a newer version is needed:

sudo apt-get remove cmake
sudo snap install cmake --classic

Issue 8: Permission Errors (when accessing dataset files)

Error:

[ERROR] Failed to access dataset path.

Fix:

Ensure the dataset files have the correct permissions:

chmod -R 755 /path/to/your/dataset

Conclusion

This guide covers the most common issues when setting up ORB-SLAM3 for event-based SLAM in ROS. Ensure all dependencies are installed and correctly linked, and always source your workspace:

source ~/catkin_ws/devel/setup.bash

If you still encounter issues, consider checking ROS logs:

roslaunch package_name node_name.launch --screen

Feel free to report any additional problems!

License

This project is open-source and follows the licensing terms of EORB_SLAM and rpg_dvs_ros.