Skip to content
BranchesTags

Latest commit

 

History

5 Commits

Folders and files

NameName
Last commit message
Last commit date
include/pointcloud_fusion
include/pointcloud_fusion
 
 
src
src
 
 
CMakeLists.txt
CMakeLists.txt
 
 
README.md
README.md
 
 
package.xml
package.xml
 
 

Repository files navigation

PointCloud Fusion

English | 中文

English

Overview

pointcloud_fusion is a ROS 2 package designed to synchronize and fuse point clouds from multiple LiDAR sensors. It uses TF2 transformations to convert point clouds to a common reference frame and merges them into a single unified point cloud.

Features

  • Multi-LiDAR Synchronization: Synchronizes point clouds from two LiDAR sensors using approximate time synchronization
  • TF2 Integration: Automatically transforms point clouds to a target reference frame using TF2
  • Point Cloud Fusion: Merges multiple point clouds into a single unified output
  • ROS 2 Native: Built specifically for ROS 2 with modern C++ practices
  • Configurable: Flexible parameter configuration for different sensor setups

Requirements

System Requirements

  • ROS 2 (tested with ROS 2 humble)
  • Ubuntu 22.04 or later (or compatible Linux distribution)
  • CMake 3.22 or later

Dependencies

  • rclcpp - ROS 2 C++ client library
  • sensor_msgs - ROS 2 sensor messages
  • message_filters - Message synchronization
  • pcl_ros - Point Cloud Library ROS interface
  • pcl_conversions - PCL/ROS conversions
  • tf2_ros - TF2 ROS bindings
  • Eigen3 - Linear algebra library

Installation

  1. Create a ROS 2 workspace (if you don't have one):

    mkdir -p ~/ros2_ws/src
cd ~/ros2_ws/src

  2. Clone the repository:

    git clone https://github.com/qza36/pointcloud_fusion.git

  3. Install dependencies:

    cd ~/ros2_ws
rosdep install --from-paths src --ignore-src -r -y

  4. Build the package:

    cd ~/ros2_ws
colcon build --packages-select pointcloud_fusion

  5. Source the workspace:

    source ~/ros2_ws/install/setup.bash

Usage

Launch the Node

ros2 run pointcloud_fusion fusion_node

With Custom Parameters

ros2 run pointcloud_fusion fusion_node --ros-args \
  -p target_frame:=base_link \
  -p lidar_topic:=/BottomLidar/livox/lidar/pointcloud \
  -p rear_lidar_topic:=/TopLidar/livox/lidar/pointcloud \
  -p lidar_frame:=livox_frame_bottom \
  -p rear_lidar_frame:=livox_frame_top

Configuration

The node supports the following ROS 2 parameters:

Parameter Type Default Value Description
target_frame string base_link Target reference frame for point cloud fusion
lidar_topic string /BottomLidar/livox/lidar/pointcloud First LiDAR point cloud topic
rear_lidar_topic string /TopLidar/livox/lidar/pointcloud Second LiDAR point cloud topic
lidar_frame string livox_frame_bottom Frame ID of the first LiDAR
rear_lidar_frame string livox_frame_top Frame ID of the second LiDAR

Topics

Subscribed Topics

  • First LiDAR point cloud (default: /BottomLidar/livox/lidar/pointcloud)

    • Type: sensor_msgs/msg/PointCloud2
    • Point cloud from the first LiDAR sensor

  • Second LiDAR point cloud (default: /TopLidar/livox/lidar/pointcloud)

    • Type: sensor_msgs/msg/PointCloud2
    • Point cloud from the second LiDAR sensor

Published Topics

  • /points/fused
    • Type: sensor_msgs/msg/PointCloud2
    • Fused point cloud in the target reference frame

How It Works

  1. Subscription: The node subscribes to two point cloud topics from different LiDAR sensors
  2. Synchronization: Uses approximate time synchronization to match point clouds from both sensors
  3. Transformation: Transforms both point clouds to the target reference frame using TF2
  4. Fusion: Concatenates the transformed point clouds into a single unified point cloud
  5. Publication: Publishes the fused point cloud to the output topic

Architecture

┌─────────────────┐
│  LiDAR Sensor 1 │
│  (Bottom)       │
└────────┬────────┘
         │ PointCloud2
         ▼
    ┌────────────────────┐
    │                    │
    │  Message Filters   │◄─────┐
    │  (Synchronizer)    │      │ PointCloud2
    │                    │      │
    └────────┬───────────┘      │
             │                  │
             │            ┌─────┴────────┐
             │            │ LiDAR Sensor 2│
             │            │ (Top)         │
             ▼            └───────────────┘
    ┌─────────────────────┐
    │  TF2 Transform      │
    │  to target_frame    │
    └────────┬────────────┘
             │
             ▼
    ┌─────────────────────┐
    │  Point Cloud Fusion │
    └────────┬────────────┘
             │
             ▼
    ┌─────────────────────┐
    │  /points/fused      │
    │  (Output)           │
    └─────────────────────┘

Development

Building for Development

cd ~/ros2_ws
colcon build --packages-select pointcloud_fusion --symlink-install

Running Tests

cd ~/ros2_ws
colcon test --packages-select pointcloud_fusion

Troubleshooting

Transform Errors

If you encounter transform errors:

  • Ensure TF2 transformations are properly broadcasted between sensor frames and the target frame
  • Check that frame IDs match between the point cloud messages and TF tree
  • Use ros2 run tf2_ros tf2_echo <source_frame> <target_frame> to verify transforms

Synchronization Issues

If point clouds are not being synchronized:

  • Verify both LiDAR topics are publishing data
  • Check that timestamps on both topics are reasonably close (within ~100ms)
  • Adjust the synchronization queue size if needed

License

TODO: License declaration

Maintainer

Contributing

Contributions are welcome! Please feel free to submit issues or pull requests.

中文

概述

pointcloud_fusion 是一个 ROS 2 软件包，用于同步和融合来自多个激光雷达传感器的点云数据。它使用 TF2 变换将点云转换到共同的参考坐标系，并将它们合并为单个统一的点云。

功能特性

  • 多激光雷达同步: 使用近似时间同步算法同步来自两个激光雷达传感器的点云
  • TF2 集成: 使用 TF2 自动将点云转换到目标参考坐标系
  • 点云融合: 将多个点云合并为单个统一输出
  • 原生 ROS 2: 专为 ROS 2 构建，采用现代 C++ 实践
  • 可配置: 灵活的参数配置，适用于不同的传感器设置

系统要求

系统环境

  • ROS 2（已在 ROS 2 humble上测试）
  • Ubuntu 22.04 或更高版本（或兼容的 Linux 发行版）
  • CMake 3.22 或更高版本

依赖项

  • rclcpp - ROS 2 C++ 客户端库
  • sensor_msgs - ROS 2 传感器消息
  • message_filters - 消息同步
  • pcl_ros - 点云库 ROS 接口
  • pcl_conversions - PCL/ROS 转换
  • tf2_ros - TF2 ROS 绑定
  • Eigen3 - 线性代数库

安装

  1. 创建 ROS 2 工作空间（如果还没有）:

    mkdir -p ~/ros2_ws/src
cd ~/ros2_ws/src

  2. 克隆仓库:

    git clone https://github.com/qza36/pointcloud_fusion.git

  3. 安装依赖项:

    cd ~/ros2_ws
rosdep install --from-paths src --ignore-src -r -y

  4. 编译软件包:

    cd ~/ros2_ws
colcon build --packages-select pointcloud_fusion

  5. 加载工作空间:

    source ~/ros2_ws/install/setup.bash

使用方法

启动节点

ros2 run pointcloud_fusion fusion_node

使用自定义参数

ros2 run pointcloud_fusion fusion_node --ros-args \
  -p target_frame:=base_link \
  -p lidar_topic:=/BottomLidar/livox/lidar/pointcloud \
  -p rear_lidar_topic:=/TopLidar/livox/lidar/pointcloud \
  -p lidar_frame:=livox_frame_bottom \
  -p rear_lidar_frame:=livox_frame_top

配置参数

节点支持以下 ROS 2 参数：

参数 类型 默认值 描述
target_frame string base_link 点云融合的目标参考坐标系
lidar_topic string /BottomLidar/livox/lidar/pointcloud 第一个激光雷达点云话题
rear_lidar_topic string /TopLidar/livox/lidar/pointcloud 第二个激光雷达点云话题
lidar_frame string livox_frame_bottom 第一个激光雷达的坐标系 ID
rear_lidar_frame string livox_frame_top 第二个激光雷达的坐标系 ID

话题

订阅话题

  • 第一个激光雷达点云（默认：/BottomLidar/livox/lidar/pointcloud）

    • 类型：sensor_msgs/msg/PointCloud2
    • 来自第一个激光雷达传感器的点云数据

  • 第二个激光雷达点云（默认：/TopLidar/livox/lidar/pointcloud）

    • 类型：sensor_msgs/msg/PointCloud2
    • 来自第二个激光雷达传感器的点云数据

发布话题

  • /points/fused
    • 类型：sensor_msgs/msg/PointCloud2
    • 目标参考坐标系中的融合点云

工作原理

  1. 订阅: 节点订阅来自不同激光雷达传感器的两个点云话题
  2. 同步: 使用近似时间同步来匹配来自两个传感器的点云
  3. 变换: 使用 TF2 将两个点云转换到目标参考坐标系
  4. 融合: 将转换后的点云连接成单个统一的点云
  5. 发布: 将融合后的点云发布到输出话题

架构图

┌─────────────────┐
│  激光雷达传感器 1 │
│  (底部)         │
└────────┬────────┘
         │ PointCloud2
         ▼
    ┌────────────────────┐
    │                    │
    │  消息过滤器         │◄─────┐
    │  (同步器)          │      │ PointCloud2
    │                    │      │
    └────────┬───────────┘      │
             │                  │
             │            ┌─────┴────────┐
             │            │ 激光雷达传感器 2│
             │            │ (顶部)        │
             ▼            └───────────────┘
    ┌─────────────────────┐
    │  TF2 坐标变换       │
    │  到 target_frame    │
    └────────┬────────────┘
             │
             ▼
    ┌─────────────────────┐
    │  点云融合            │
    └────────┬────────────┘
             │
             ▼
    ┌─────────────────────┐
    │  /points/fused      │
    │  (输出)             │
    └─────────────────────┘

开发

开发构建

cd ~/ros2_ws
colcon build --packages-select pointcloud_fusion --symlink-install

运行测试

cd ~/ros2_ws
colcon test --packages-select pointcloud_fusion

故障排除

变换错误

如果遇到变换错误：

  • 确保在传感器坐标系和目标坐标系之间正确广播了 TF2 变换
  • 检查点云消息的坐标系 ID 是否与 TF 树匹配
  • 使用 ros2 run tf2_ros tf2_echo <source_frame> <target_frame> 验证变换

同步问题

如果点云无法同步：

  • 验证两个激光雷达话题是否正在发布数据
  • 检查两个话题的时间戳是否足够接近（在 ~100ms 以内）
  • 如有必要，调整同步队列大小

许可证

待定：许可证声明

维护者

贡献

欢迎贡献！请随时提交问题或拉取请求。

About

No description, website, or topics provided.

Resources

Readme
Activity

Stars

0 stars

Watchers

0 watching

Forks

0 forks
Report repository

Releases

Packages

Contributors

Languages