CAPO-LeggedRobotOdometry 🦾

Language / 语言： English | 中文

CAPO-LeggedRobotOdometry is a ROS 2 (Humble) proprioceptive odometry / state-estimation repository for biped / quadruped / wheel-legged robots on Ubuntu 22.04.

It provides high-accuracy odometry using only IMU + joint encoders + foot contact/force signals, without requiring cameras or LiDAR.

In 3D closed-loop trials (a 200 m horizontal and 15 m vertical loop), Astrall point-foot robot A achieves 0.1638 m horizontal error and 0.219 m vertical error; for wheel-legged robot B, the corresponding errors are 0.2264 m and 0.199 m.

At the repository level, fusion_estimator_node.cpp is mainly the ROS2 wrapper (topics, parameters, message conversion, publishing), while the actual fusion-estimation algorithm lives in FusionEstimator/ as a portable, pure C++ implementation. This makes it straightforward to port the estimator to ROS1, non-ROS2 applications, or embedded platforms.

In addition, the Matlab/ folder provides a MATLAB + MEX example that compiles and calls the same C++ estimator core for offline validation, algorithm analysis, and cross-platform reuse.

📄 Paper

Contact-Anchored Proprioceptive Odometry for Quadruped Robots (arXiv:2602.17393)

Paper: https://arxiv.org/abs/2602.17393

If you use this repository in research, please consider citing the paper.

📦 Data Sharing (Go2-EDU ROS bags)

To help readers quickly validate the pipeline, we provide two Go2-EDU trial datasets, including real-world videos and the corresponding ROS bag topics/messages required by this node, enabling fast reproduction and sanity checks.

Download (Google Drive): https://drive.google.com/drive/folders/1FfVO69rfmUu6B9crPhZCfKf9wFnV4L7n?usp=sharing

Note: the IMU on this Go2-EDU platform exhibits noticeable yaw drift, so the odometry accuracy is generally worse than the results reported for Astrall robots A and B in the paper.

✨ Key Features

Category	Description
Biped / Quadruped / Wheel-Legged Unified	Online contact-set switching; stance legs are detected automatically, supporting fast transitions between standing and walking.
Full 3D & Planar 2D	Publishes both 6DoF odometry (`SMX/Odom`) and a gravity-flattened 2D odometry (`SMX/Odom_2D`).
No Exteroception Required	Works without cameras or LiDAR; only IMU, joint encoders, and foot contact/force signals are required.
Portable Pure C++ Core	The estimator core is isolated in `FusionEstimator/`, making it easier to reuse outside ROS2.
MATLAB / MEX Validation	The `Matlab/` folder demonstrates how to compile and invoke the same C++ core from MATLAB.
Runtime Tuning	Key parameters can be adjusted through `config.yaml`, and platform-dependent thresholds can be tuned for different robots.

🏗️ Related Repositories

Scope	Repository	Summary
Low-level / Drivers	https://github.com/ShineMinxing/Ros2Go2Base	DDS bridge, Unitree SDK2 control, pointcloud→LaserScan, TF utilities
Odometry	CAPO-LeggedRobotOdometry (this repo)	Pure proprioceptive fusion, publishes `SMX/Odom` / `SMX/Odom_2D`
SLAM / Mapping	https://github.com/ShineMinxing/Ros2SLAM	Integrations for Cartographer 3D, KISS-ICP, FAST-LIO2, Point-LIO
Voice / LLM	https://github.com/ShineMinxing/Ros2Chat	Offline ASR + OpenAI Chat + TTS
Vision	https://github.com/ShineMinxing/Ros2ImageProcess	Camera pipelines, spot / face / drone detection
Gimbal Tracking	https://github.com/ShineMinxing/Ros2AmovG1	Amov G1 gimbal control and tracking
Tools	https://github.com/ShineMinxing/Ros2Tools	Bluetooth IMU, joystick mapping, gimbal loop, data logging

⚠️ Clone as needed. If you only need state estimation, this repository is sufficient. For mapping, it is natural to pair it with Ros2SLAM and Ros2Go2Base.

📂 Repository Layout

CAPO-LeggedRobotOdometry/
├── CMakeLists.txt
├── package.xml
├── config.yaml
├── fusion_estimator_node.cpp        # ROS2 node wrapper: params / topics / odom publishing
├── FusionEstimator/                 # pure C++ fusion-estimation core (portable)
│   ├── Estimators/
│   ├── fusion_estimator.h
│   ├── LowlevelState.h
│   ├── SensorBase.cpp
│   ├── SensorBase.h
│   ├── Sensor_IMU.cpp
│   ├── Sensor_IMU.h
│   ├── Sensor_Legs.cpp
│   ├── Sensor_Legs.h
│   └── Readme.md
├── Matlab/                          # MATLAB + MEX example for calling the C++ core
│   ├── build_mex.m
│   ├── fusion_estimator.m
│   ├── fusion_estimator_mex.cpp
│   ├── MPXY150Z10.csv
│   └── MWXY150Z10.csv
├── Plotjuggler.xml                  # PlotJuggler layout / visualization helper
└── Readme.md

🧩 Architecture Notes

This repository is intentionally split into two layers:

1. ROS2 wrapper layer

fusion_estimator_node.cpp is responsible for:

reading parameters from config.yaml
subscribing to ROS2 topics
converting ROS messages into estimator inputs
invoking the fusion-estimation core
publishing nav_msgs/Odometry

2. Portable estimator core

FusionEstimator/ contains the actual estimation logic:

IMU handling
leg kinematics / leg sensors
contact-anchored fusion logic
estimator state update
reusable C++ interfaces

Because this part is pure C++, it can be reused in:

ROS1
non-ROS Linux applications
embedded / custom runtime systems
MATLAB via MEX

3. MATLAB / MEX bridge

Matlab/ is not required for ROS2 runtime, but it is useful when you want to:

validate the estimator offline
replay data in MATLAB
compare C++ results with MATLAB scripts
reuse the same FusionEstimator/ core without ROS2

⚙️ Configuration (`config.yaml`)

The table below lists the main parameters used by the ROS2 node. See config.yaml for the full file and comments.

Parameter	Typical Value	Meaning
`sub_imu_topic`	`SMX/Go2IMU`	IMU topic
`sub_joint_topic`	`SMX/Go2Joint`	joint state topic
`sub_mode_topic`	`SMX/SportCmd`	reset / mode topic
`pub_odom_topic`	`SMX/Odom`	6DoF odometry output
`pub_odom2d_topic`	`SMX/Odom_2D`	planar odometry output
`odom_frame`	`odom`	odometry frame
`base_frame`	`base_link`	body frame
`base_frame_2d`	`base_link_2D`	planar body frame
`imu_data_enable`	`true`	enable IMU input
`leg_pos_enable`	`true`	enable leg-position kinematics
`leg_vel_enable`	`true`	enable leg-velocity kinematics
`leg_ori_enable`	`false`	enable kinematics-based yaw correction
`leg_ori_init_weight`	`0.001`	initial weight of orientation correction
`leg_ori_time_wight`	`1000.0`	time-related growth weight for orientation correction
`leg_velcke_enable`	`false`	enable IKVel-CKF / CAPO-CKE
`foot_force_threshold`	`20.0`	contact threshold
`min_stair_height`	`0.08`	minimum stair-height hypothesis used by the estimator

Notes:

The exact values are platform-dependent.

Some fields in config.yaml may be experimental, legacy, or intended for specific branches / workflows.

If you port only the pure C++ core, you usually need to keep only the estimator-relevant fields and can drop ROS2-specific naming.

⚙️ Build & Run

Clone the repository

mkdir -p ~/ros2_ws/src
cd ~/ros2_ws/src
git clone --depth 1 https://github.com/ShineMinxing/CAPO-LeggedRobotOdometry.git
cd ~/ros2_ws
colcon build --packages-select fusion_estimator
source install/setup.bash
ros2 run fusion_estimator fusion_estimator_node

Optional: pull large files / sample data

If you want to use optional sample CSV files or large assets, you may need Git LFS:

sudo apt update
sudo apt install git-lfs
git lfs install
cd ~/ros2_ws/src/CAPO-LeggedRobotOdometry
git lfs pull

Note: in some versions of this repository, the node loads config.yaml through a hard-coded --params-file path inside fusion_estimator_node.cpp. If your local workspace path differs, update that path and rebuild.

📑 ROS 2 Interfaces

/fusion_estimator_node (rclcpp)
├─ Publishes
│   • SMX/Odom         nav_msgs/Odometry (odom → base_link)
│   • SMX/Odom_2D      nav_msgs/Odometry (odom → base_link_2D)
├─ Subscribes
│   • SMX/Go2IMU       sensor_msgs/Imu
│   • SMX/Go2Joint     std_msgs/Float64MultiArray
│       layout:
│         data[0..11]   = 12 joint positions q
│         data[12..23]  = 12 joint velocities dq
│         data[24..27]  = 4 foot-force / contact-related values
│   • SMX/SportCmd     std_msgs/Float64MultiArray
│       reset example:
│         data[0] == 25140000  → estimator reset
└─ TF is typically published by another node / utility if needed

🧠 Algorithm Overview (High Level)

Contact detection
Detect stance legs from force / threshold logic.
Forward kinematics
Compute foot-end position / velocity in the body frame from joint measurements.
Contact anchoring
Record touchdown footfall points and use them as intermittent world-frame constraints during stance.
Height stabilization
Use support-plane height logic and confidence decay to reduce long-horizon elevation drift.
Optional IKVel-CKF (CAPO-CKE)
Suppress encoder-induced velocity spikes and obtain smoother leg-end velocity estimates.
Optional yaw stabilization
Use multi-contact geometric consistency to reduce IMU yaw drift during prolonged standing.
2D projection
Publish a planar odometry (SMX/Odom_2D) by flattening roll / pitch while preserving yaw and planar motion.

For full derivations, please refer to the paper.

🔧 Reusing `FusionEstimator/` Outside ROS2

If you do not need ROS2, the reusable part is mainly under FusionEstimator/.

A typical migration path is:

keep the estimator core in FusionEstimator/
replace ROS2 message subscriptions with your own sensor interface
prepare IMU, joint position / velocity, and contact / force inputs
call the estimator core from your own loop
export odometry to your target middleware / runtime

This design is useful for:

ROS1 migration
embedded deployment
offline replay tools
MATLAB / simulation analysis
custom robotics middleware

🧪 MATLAB / MEX (Optional)

The Matlab/ folder provides an additional example for compiling the C++ estimator core into a MATLAB MEX module.

Typical workflow:

cd Matlab
build_mex
fusion_estimator

Files in this folder:

build_mex.m — build script for MEX compilation
fusion_estimator_mex.cpp — MEX bridge wrapping the C++ estimator core
fusion_estimator.m — MATLAB-side usage / validation example
MPXY150Z10.csv, MWXY150Z10.csv — sample data for offline tests

This is especially useful when you want to:

verify C++ estimator behavior in MATLAB
compare offline results quickly
reuse the same FusionEstimator/ implementation without ROS2
debug algorithm changes before reintegrating them into the ROS2 node

📈 Visualization

A Plotjuggler.xml file is included in the repository for convenient visualization and debugging with PlotJuggler.

You can use it to inspect:

odometry outputs
IMU-related signals
joint / force-related channels
estimator behavior during walking, standing, or reset events

🎥 Videos

Topic	Link
Odometry-only mapping (morphology switching)
Indoor walking (0.5%–1% error)
Stair climbing (height error < 5 cm)
Outdoor walking (380 m, 3.3% error)
Voice interaction + navigation
Face tracking + laser spot tracking
AR glasses head-following
YOLO drone tracking
Gimbal + fixed camera collaboration
Multiple SLAM integrations

📄 Further Reading

Paper (arXiv): https://arxiv.org/abs/2602.17393
Technical notes (Notion): https://www.notion.so/Ros2Go2-1e3a3ea29e778044a4c9c35df4c27b22
ROS1 reference implementation: https://github.com/ShineMinxing/FusionEstimation

📨 Contact

Email	Affiliation
sunminxing20@mails.ucas.ac.cn	Institute of Optics and Electronics, CAS

This repository is under active development. Issues and PRs are welcome.

CAPO-LeggedRobotOdometry 🦾 (中文)

语言 / Language： English | 中文

CAPO-LeggedRobotOdometry 是面向 双足 / 四足 / 轮足 机器人的 ROS 2（Humble） 本体状态估计 / 里程计算法仓库，可在 Ubuntu 22.04 上运行。

本仓库仅依赖 IMU + 关节编码器 + 足端接触/力信号 即可输出高精度里程计，不依赖相机或激光雷达。

在 3D 闭环运动实验（水平 200 m、竖直 15 m）中：Astrall 公司的点足机器人 A 的水平与竖直误差分别为 0.1638 m 和 0.219 m；轮足机器人 B 的对应误差分别为 0.2264 m 和 0.199 m。

从工程结构上看，fusion_estimator_node.cpp 主要负责 ROS2 封装层（参数、Topic、消息转换、发布），而真正的融合估计算法位于 FusionEstimator/ 中，并且实现为 纯 C++、可移植 的核心模块。因此它不仅能用于 ROS2，也便于迁移到：

ROS1
非 ROS2 平台
嵌入式 / 自定义运行环境

另外，仓库中的 Matlab/ 文件夹额外提供了 MATLAB + MEX 混合编译并调用 FusionEstimator 中 C++ 核心代码 的示例，方便离线验证、算法分析和跨平台复用。

📄 论文

Contact-Anchored Proprioceptive Odometry for Quadruped Robots（arXiv:2602.17393）

论文链接：https://arxiv.org/abs/2602.17393

如果本仓库对你的研究有帮助，欢迎引用论文。

📦 数据共享

为便于大家快速验证本仓库的有效性，我们提供了 Go2-EDU 的两段示例数据，包含实拍录像以及可直接用于本节点的 ROS bag topic / message，用于快速复现与 sanity check。

数据下载（Google Drive）：https://drive.google.com/drive/folders/1FfVO69rfmUu6B9crPhZCfKf9wFnV4L7n?usp=sharing

说明：该 Go2-EDU 平台的 IMU yaw 漂移较明显，因此里程计精度通常会劣于论文中 Astrall 机器人 A 与 B 的实验结果。

✨ 算法特性

类别	说明
双足 / 四足 / 轮足统一支持	运行中自动识别支撑足，支持站立、行走和形态切换。
全 3D + 平面 2D	同时发布 6DoF 里程计 `SMX/Odom` 与压平后的 `SMX/Odom_2D`。
零依赖外部感知	不依赖相机或激光雷达，仅依赖本体传感器。
纯 C++ 可移植核心	`FusionEstimator/` 中是独立的纯 C++ 融合估计实现，便于脱离 ROS2 使用。
MATLAB / MEX 联调	`Matlab/` 中给出了用 MATLAB 混编调用同一套 C++ 核心的示例。
参数可调	`config.yaml` 中的关键参数和阈值可按平台进行调整。

🏗️ 生态仓库一览

范畴	仓库	功能简介
底层驱动	https://github.com/ShineMinxing/Ros2Go2Base	DDS 桥、Unitree SDK2 控制、点云→Scan、TF 工具
里程计	CAPO-LeggedRobotOdometry（本仓库）	纯本体多传感器融合，发布 `SMX/Odom` / `SMX/Odom_2D`
SLAM / 建图	https://github.com/ShineMinxing/Ros2SLAM	集成 Cartographer 3D、KISS-ICP、FAST-LIO2、Point-LIO 等
语音 / LLM	https://github.com/ShineMinxing/Ros2Chat	离线 ASR + OpenAI Chat + 语音合成
图像处理	https://github.com/ShineMinxing/Ros2ImageProcess	相机、光点 / 人脸 / 无人机检测
吊舱跟随	https://github.com/ShineMinxing/Ros2AmovG1	Amov G1 吊舱控制与目标跟踪
工具集	https://github.com/ShineMinxing/Ros2Tools	蓝牙 IMU、手柄映射、吊舱闭环、数据记录等

⚠️ 按需克隆：如果只做状态估计，本仓库即可独立使用；如果要建图，通常可以配合 Ros2SLAM 和 Ros2Go2Base 一起使用。

📂 仓库结构

CAPO-LeggedRobotOdometry/
├── CMakeLists.txt
├── package.xml
├── config.yaml
├── fusion_estimator_node.cpp        # ROS2 节点封装：参数 / Topic / 里程计发布
├── FusionEstimator/                 # 纯 C++ 融合估计算法核心（可移植）
│   ├── Estimators/
│   ├── fusion_estimator.h
│   ├── LowlevelState.h
│   ├── SensorBase.cpp
│   ├── SensorBase.h
│   ├── Sensor_IMU.cpp
│   ├── Sensor_IMU.h
│   ├── Sensor_Legs.cpp
│   ├── Sensor_Legs.h
│   └── Readme.md
├── Matlab/                          # MATLAB + MEX 调用 C++ 核心的示例
│   ├── build_mex.m
│   ├── fusion_estimator.m
│   ├── fusion_estimator_mex.cpp
│   ├── MPXY150Z10.csv
│   └── MWXY150Z10.csv
├── Plotjuggler.xml                  # PlotJuggler 可视化配置
└── Readme.md

🧩 架构说明

本仓库有意分成两层：

1）ROS2 封装层

fusion_estimator_node.cpp 主要负责：

从 config.yaml 读取参数
订阅 ROS2 Topic
将 ROS 消息转换为估计器输入
调用融合估计算法核心
发布 nav_msgs/Odometry

2）可移植的估计器核心

FusionEstimator/ 中包含真正的估计逻辑，例如：

IMU 数据处理
腿部运动学 / 腿部传感器建模
接触锚定约束
融合估计状态更新
可复用的 C++ 接口

因为这里是纯 C++ 实现，所以它不仅能在 ROS2 中使用，也很适合迁移到：

ROS1
非 ROS2 的 Linux 程序
嵌入式 / 自定义系统
MATLAB MEX
离线回放与仿真验证

3）MATLAB / MEX 桥接层

Matlab/ 文件夹不是 ROS2 运行必须的，但在以下场景里很有用：

做离线数据验证
在 MATLAB 中回放与分析估计结果
快速对比 C++ 结果与 MATLAB 脚本
不依赖 ROS2 直接复用同一套 FusionEstimator/ 核心

⚙️ 参数配置（`config.yaml`）

下面列出当前 ROS2 节点最常用的参数。完整注释请查看 config.yaml。

参数名	常见取值	说明
`sub_imu_topic`	`SMX/Go2IMU`	IMU 订阅 Topic
`sub_joint_topic`	`SMX/Go2Joint`	关节状态订阅 Topic
`sub_mode_topic`	`SMX/SportCmd`	复位 / 模式 Topic
`pub_odom_topic`	`SMX/Odom`	全 3D 里程计输出
`pub_odom2d_topic`	`SMX/Odom_2D`	平面里程计输出
`odom_frame`	`odom`	里程计坐标系
`base_frame`	`base_link`	机体坐标系
`base_frame_2d`	`base_link_2D`	平面机体坐标系
`imu_data_enable`	`true`	是否启用 IMU
`leg_pos_enable`	`true`	是否启用腿部位置运动学
`leg_vel_enable`	`true`	是否启用腿部速度运动学
`leg_ori_enable`	`false`	是否启用基于运动学的航向修正
`leg_ori_init_weight`	`0.001`	姿态修正初始权重
`leg_ori_time_wight`	`1000.0`	姿态修正随时间增长的权重
`leg_velcke_enable`	`false`	是否启用 IKVel-CKF / CAPO-CKE
`foot_force_threshold`	`20.0`	触地阈值
`min_stair_height`	`0.08`	最小台阶高度假设

说明：

具体参数需要根据机器人平台做调整。

config.yaml 中有些字段可能用于实验、兼容旧版本，或特定分支。

如果你只移植 FusionEstimator/ 纯 C++ 核心，一般只需要保留估计器真正需要的参数，不必保留 ROS2 命名相关字段。

⚙️ 编译与运行

克隆仓库

mkdir -p ~/ros2_ws/src
cd ~/ros2_ws/src
git clone --depth 1 https://github.com/ShineMinxing/CAPO-LeggedRobotOdometry.git
cd ~/ros2_ws
colcon build --packages-select fusion_estimator
source install/setup.bash
ros2 run fusion_estimator fusion_estimator_node

可选：拉取大文件 / 示例数据

如果你需要示例 CSV 或大文件资源，可以安装 Git LFS：

sudo apt update
sudo apt install git-lfs
git lfs install
cd ~/ros2_ws/src/CAPO-LeggedRobotOdometry
git lfs pull

说明：某些版本中，fusion_estimator_node.cpp 会通过硬编码的 --params-file 路径加载 config.yaml。如果你的本地工作空间路径不同，需要先修改该路径并重新编译。

📑 节点接口

/fusion_estimator_node (rclcpp)
├─ 发布
│   • SMX/Odom         nav_msgs/Odometry (odom → base_link)
│   • SMX/Odom_2D      nav_msgs/Odometry (odom → base_link_2D)
├─ 订阅
│   • SMX/Go2IMU       sensor_msgs/Imu
│   • SMX/Go2Joint     std_msgs/Float64MultiArray
│       数据布局：
│         data[0..11]   = 12 个关节位置 q
│         data[12..23]  = 12 个关节速度 dq
│         data[24..27]  = 4 个足端力 / 接触相关数值
│   • SMX/SportCmd     std_msgs/Float64MultiArray
│       复位示例：
│         data[0] == 25140000  → 复位估计器
└─ 若需要 TF，一般由其他节点 / 工具单独发布

🧠 算法概览（简述）

触地检测
基于足端力 / 阈值逻辑判定支撑腿。
前向运动学
根据关节测量计算足端在机身坐标系中的位置与速度。
落足点锚定
在触地时记录落足点，在支撑期将接触点作为世界系间歇约束。
高度稳定
利用支撑平面高度逻辑与置信度衰减来降低长时程高度漂移。
可选 IKVel-CKF（CAPO-CKE）
抑制编码器速度尖峰，使足端速度估计更平滑。
可选航向稳定
利用多接触几何一致性来减小原地站立时的 IMU yaw 漂移。
2D 投影
压平 roll / pitch，保留 yaw 与平面运动，发布 SMX/Odom_2D。

完整推导请参考论文。

🔧 在 ROS2 之外复用 `FusionEstimator/`

如果你不需要 ROS2，真正可复用的主体主要就是 FusionEstimator/。

一个典型迁移流程是：

保留 FusionEstimator/ 下的纯 C++ 核心
用你自己的传感器接口替换 ROS2 订阅
准备 IMU、关节位置 / 速度、触地 / 足端力输入
在你的主循环中调用估计器
将输出接到你的中间件 / 控制系统

这对以下场景很有帮助：

ROS1 迁移
嵌入式部署
离线数据回放
MATLAB / 仿真验证
自定义机器人软件框架

🧪 MATLAB / MEX（可选）

Matlab/ 文件夹提供了一个额外示例，用来把 FusionEstimator/ 中的纯 C++ 核心编译成 MATLAB 可调用的 MEX 模块。

典型流程如下：

cd Matlab
build_mex
fusion_estimator

该文件夹中的主要文件：

build_mex.m：MEX 编译脚本
fusion_estimator_mex.cpp：MEX 桥接文件，负责封装 C++ 估计器核心
fusion_estimator.m：MATLAB 侧调用示例
MPXY150Z10.csv、MWXY150Z10.csv：离线测试用示例数据

这部分特别适合以下需求：

在 MATLAB 中验证 C++ 融合器行为
快速做离线结果对比
不依赖 ROS2 复用同一套 FusionEstimator/
在把修改重新合回 ROS2 节点之前，先做算法调试

📈 可视化

仓库中附带了 Plotjuggler.xml，可用于 PlotJuggler 的快速可视化与调试。

你可以用它观察：

里程计输出
IMU 相关信号
关节 / 足端力相关量
站立、行走、复位过程中的估计器行为

🎥 视频演示

主题	链接
纯里程计建图（形态切换）
室内行走（误差 0.5%–1%）
爬楼梯（高度误差 < 5 cm）
户外行走（380 m，误差 3.3%）
语音交互 + 地图导航
人脸识别跟踪 + 光点跟踪
AR 眼镜头部运动跟随
YOLO 无人机识别与跟随
吊舱 + 固定相机协同
多种 SLAM 方法集成

📄 延伸阅读

论文（arXiv）：https://arxiv.org/abs/2602.17393
技术笔记（Notion）：https://www.notion.so/Ros2Go2-1e3a3ea29e778044a4c9c35df4c27b22
ROS1 参考实现：https://github.com/ShineMinxing/FusionEstimation

📨 联系方式

邮箱	单位
sunminxing20@mails.ucas.ac.cn	中国科学院光电技术研究所

本仓库持续开发中，欢迎 Issue / PR 交流与贡献。

Name		Name	Last commit message	Last commit date
Latest commit History 144 Commits
FusionEstimator		FusionEstimator
Matlab		Matlab
CMakeLists.txt		CMakeLists.txt
Plotjuggler.xml		Plotjuggler.xml
README.md		README.md
config.yaml		config.yaml
fusion_estimator_node.cpp		fusion_estimator_node.cpp
package.xml		package.xml

Folders and files

Latest commit

History

Repository files navigation

CAPO-LeggedRobotOdometry 🦾

📄 Paper

📦 Data Sharing (Go2-EDU ROS bags)

✨ Key Features

🏗️ Related Repositories

📂 Repository Layout

🧩 Architecture Notes

1. ROS2 wrapper layer

2. Portable estimator core

3. MATLAB / MEX bridge

⚙️ Configuration (config.yaml)

⚙️ Build & Run

Clone the repository

Optional: pull large files / sample data

📑 ROS 2 Interfaces

🧠 Algorithm Overview (High Level)

🔧 Reusing FusionEstimator/ Outside ROS2

🧪 MATLAB / MEX (Optional)

📈 Visualization

🎥 Videos

📄 Further Reading

📨 Contact

CAPO-LeggedRobotOdometry 🦾 (中文)

📄 论文

📦 数据共享

✨ 算法特性

🏗️ 生态仓库一览

📂 仓库结构

🧩 架构说明

1）ROS2 封装层

2）可移植的估计器核心

3）MATLAB / MEX 桥接层

⚙️ 参数配置（config.yaml）

⚙️ 编译与运行

克隆仓库

可选：拉取大文件 / 示例数据

📑 节点接口

🧠 算法概览（简述）

🔧 在 ROS2 之外复用 FusionEstimator/

🧪 MATLAB / MEX（可选）

📈 可视化

🎥 视频演示

📄 延伸阅读

📨 联系方式

About

Resources

Uh oh!

Stars

Watchers

Forks

Releases

Packages 0

Uh oh!

Uh oh!

Contributors

Uh oh!

Languages

⚙️ Configuration (`config.yaml`)

🔧 Reusing `FusionEstimator/` Outside ROS2

⚙️ 参数配置（`config.yaml`）

🔧 在 ROS2 之外复用 `FusionEstimator/`

Packages