Depth Camera Simulation: RealSense D435 and Xema-S

Disclaimer

Note: Due to the lack of official Gazebo Ignition plugins for RealSense D435 and Xema-S cameras, this simulation uses Gazebo's standard RGB-D camera sensors configured to mimic the specific properties of these depth cameras. The simulation leverages ros_gz_bridge for ROS 2 integration.

Project Overview

This project provides a detailed simulation of two depth cameras:

1. Intel RealSense D435
2. Xema-S Depth Camera

The simulation includes:

• Realistic camera parameter configurations
• RGB and depth image generation
• ROS 2 topic bridging
• Mounting on a simple differential drive robot

Prerequisites

System Requirements

• Ubuntu 22.04 LTS
• ROS 2 Humble
• Gazebo Ignition Fortress

Initial Setup

1. Update System

sudo apt update && sudo apt upgrade -y

2. Install ROS 2 Humble

sudo apt install -y ros-humble-desktop

3. Install Gazebo Ignition Fortress

sudo wget https://packages.osrfoundation.org/gazebo.gpg -O /usr/share/keyrings/pkgs-osrf-archive.gpg
echo "deb [arch=amd64 signed-by=/usr/share/keyrings/pkgs-osrf-archive.gpg] https://packages.osrfoundation.org/gazebo/ubuntu-stable `lsb_release -cs` main" | sudo tee /etc/apt/sources.list.d/gazebo-stable.list
sudo apt update
sudo apt install -y ignition-fortress

4. Install ROS-Gazebo Bridge

sudo apt install -y ros-humble-ros-gz-bridge

Intel RealSense D435 Simulation

Workspace Preparation

1. Clone Repository

git clone https://github.com/aaqibmahamood/Depth_Camera_Simulation.git

3. Copy to Home Directory

# Ensure you're in the Home directory where you have your workspaces. Copy d435_ws into home from Depth_Camera_Simulation folder.
cp -r d435_ws ~/

Camera Specifications - RealSense D435

RGB Camera

• Resolution: 1920x1080
• Horizontal FOV: 80 degrees
• Update Rate: 30 Hz
• Distortion Model: Realistic radial and tangential distortion
• Intrinsic Camera Matrix: Calibrated to D435 specifications

Depth Camera

• Resolution: 1280x720
• Horizontal FOV: 80 degrees
• Depth Range: 0.2m - 10m
• Noise Model: Gaussian noise (mean 0, stddev 0.0025)
• Update Rate: 30 Hz

Building and Launching

1. Build Workspace

cd ~/d435_ws
colcon build
source install/setup.bash

2. Launch Simulation

ros2 launch depth_d435 one_robot_ign_launch.py

Visualization Examples

Intel RealSense D435 DAE File

Demonstration Video

d435-simulation.mp4

Xema-S Depth Camera Simulation

Workspace Preparation

1. Clone Repository

git clone <repository-url-for-xema-s>

2. Copy to Home Directory

# Ensure you're in the src directory
cp -r xema_ws ~/

Camera Specifications - Xema-S

RGB Camera

• Resolution: 1920x1080
• Horizontal FOV: 70 degrees
• Update Rate: 30 Hz
• Distortion Coefficients: Precisely tuned
• Intrinsic Camera Matrix: Matched to Xema-S optics

Depth Camera

• Resolution: 1280x720
• Horizontal FOV: 70 degrees
• Depth Range: 0.1m - 8m
• Noise Model: Gaussian noise (mean 0, stddev 0.0025)
• Update Rate: 30 Hz

Building and Launching

1. Build Workspace

cd ~/xema_ws
colcon build
source install/setup.bash

2. Launch Simulation

ros2 launch xema_s one_robot_ign_launch.py

Visualization Examples

Open3D's XEMA-S DAE File

Advanced Configuration

Optical Frames

• Dedicated optical frames for both RGB and depth cameras
• Precise alignment with base link
• Consistent orientation and positioning

Simulation Parameters

• Clip Distances:
  • D435: Near clip 0.2m, Far clip 10m
  • Xema-S: Near clip 0.1m, Far clip 15m
• Accurate lens intrinsics and distortion parameters

Bridging ROS 2 Topics

To bridge camera topics:

ros2 run ros_gz_bridge parameter_bridge /camera/depth/image@sensor_msgs/msg/Image@ignition.msgs.Image
ros2 run ros_gz_bridge parameter_bridge /camera/color/image@sensor_msgs/msg/Image@ignition.msgs.Image

Troubleshooting

• Verify ROS 2 and Gazebo Ignition installations
• Check topic names and message types
• Ensure correct camera placement in Gazebo world

Contributing

Contributions are welcome! 🤖

1. Fork the repository
2. Create your feature branch
3. Commit your changes
4. Push to the branch
5. Open a Pull Request

Acknowledgments

• ROS 2 Community
• Gazebo Ignition Team
• Intel RealSense
• Xema-S Manufacturers