cartographer

Cartographer is a system that provides real-time simultaneous localization and mapping (SLAM) in 2D and 3D across multiple platforms and sensor configurations.

7,089

2,247

7,089

231

View on GitHub

Top Related Projects

cartographer

7,090

Cartographer is a system that provides real-time simultaneous localization and mapping (SLAM) in 2D and 3D across multiple platforms and sensor configurations.

rtabmap

2,713

RTAB-Map library and standalone application

ORB_SLAM2

9,286

Real-Time SLAM for Monocular, Stereo and RGB-D Cameras, with Loop Detection and Relocalization Capabilities

Quick Overview

Cartographer is an open-source system for real-time simultaneous localization and mapping (SLAM) in 2D and 3D across multiple platforms and sensor configurations. It is developed by Google and provides a flexible and efficient solution for robotics and autonomous systems to create maps of their environment while simultaneously determining their location within it.

Pros

High performance and real-time operation
Supports both 2D and 3D SLAM
Flexible and adaptable to various sensor configurations
Well-documented and actively maintained by Google

Cons

Steep learning curve for beginners
Requires careful tuning of parameters for optimal performance
Limited support for loop closure in certain scenarios
Resource-intensive for large-scale mapping tasks

Code Examples

Initializing a 2D SLAM node:

#include "cartographer/mapping/map_builder.h"
#include "cartographer/sensor/range_data.h"

auto map_builder = cartographer::mapping::CreateMapBuilder(options);
auto trajectory_builder = map_builder->AddTrajectoryBuilder(
    {sensor_id}, trajectory_options, nullptr);

Adding sensor data to the SLAM system:

sensor::TimedPointCloudData point_cloud;
// ... populate point_cloud with sensor data
trajectory_builder->AddSensorData(sensor_id, point_cloud);

Retrieving the current pose estimate:

auto pose_estimate = trajectory_builder->pose_estimate();
if (pose_estimate.has_value()) {
    const auto& pose = pose_estimate.value().pose;
    std::cout << "Current pose: " << pose << std::endl;
}

Getting Started

To get started with Cartographer:

Install dependencies:

sudo apt-get install -y google-mock libboost-all-dev libeigen3-dev libgflags-dev libgoogle-glog-dev liblua5.2-dev libprotobuf-dev libsuitesparse-dev libwebp-dev ninja-build protobuf-compiler python-sphinx

Clone and build Cartographer:

git clone https://github.com/cartographer-project/cartographer.git
cd cartographer && mkdir build && cd build
cmake .. -G Ninja
ninja
ninja test
sudo ninja install

Run a demo:

cartographer_assets_writer --configuration_directory ~/cartographer/configuration_files --configuration_basename assets_writer_backpack_2d.lua --urdf_filename ~/cartographer/configuration_files/backpack_2d.urdf --bag_filenames ~/cartographer_ros/cartographer_ros/cartographer_ros/test_data/backpack_2d.bag --output_file_prefix ~/cartographer_ros/cartographer_ros/test_data/backpack_2d

Competitor Comparisons

cartographer

7,090

Cartographer is a system that provides real-time simultaneous localization and mapping (SLAM) in 2D and 3D across multiple platforms and sensor configurations.

Pros of Cartographer

More active development and frequent updates
Larger community and contributor base
Broader range of supported platforms and sensors

Cons of Cartographer

Higher complexity and steeper learning curve
Potentially higher computational requirements
Less flexibility for customization in certain scenarios

Code Comparison

Cartographer:

void LocalTrajectoryBuilder2D::AddImuData(const sensor::ImuData& imu_data) {
  CHECK(timed_pose_queue_.empty() ||
        imu_data.time >= timed_pose_queue_.back().time);
  imu_data_.push_back(imu_data);
  TrimImuData();
}

Cartographer>:

void LocalTrajectoryBuilder2D::AddImuData(const sensor::ImuData& imu_data) {
  CHECK(timed_pose_queue_.empty() ||
        imu_data.time >= timed_pose_queue_.back().time);
  imu_data_.push_back(imu_data);
}

The main difference in this code snippet is that Cartographer includes a TrimImuData() function call, which is not present in Cartographer>. This suggests that Cartographer may have additional data management features or optimizations in place.

rtabmap

2,713

RTAB-Map library and standalone application

Pros of RTAB-Map

More versatile, supporting various sensors and input types (stereo, RGB-D, lidar)
Better suited for large-scale mapping and loop closure detection
Includes built-in visualization tools and GUI for easier data interpretation

Cons of RTAB-Map

Generally slower processing speed compared to Cartographer
May require more computational resources, especially for large maps
Steeper learning curve due to more complex configuration options

Code Comparison

RTAB-Map (C++):

#include <rtabmap/core/Rtabmap.h>
#include <rtabmap/core/RtabmapThread.h>

rtabmap::Rtabmap rtabmap;
rtabmap::RtabmapThread rtabmapThread(&rtabmap);
rtabmapThread.start();

Cartographer (C++):

#include "cartographer/mapping/map_builder.h"
#include "cartographer/sensor/range_data.h"

auto map_builder = cartographer::mapping::CreateMapBuilder(map_builder_options);
map_builder->AddTrajectoryBuilder(trajectory_options);

Both libraries offer powerful SLAM capabilities, but RTAB-Map is more versatile and feature-rich, while Cartographer focuses on efficiency and real-time performance. The choice between them depends on specific project requirements and hardware constraints.

ORB_SLAM2

9,286

Real-Time SLAM for Monocular, Stereo and RGB-D Cameras, with Loop Detection and Relocalization Capabilities

Pros of ORB_SLAM2

Lightweight and efficient, suitable for real-time applications
Robust loop closure and relocalization capabilities
Works well in feature-rich environments

Cons of ORB_SLAM2

Limited to visual SLAM, doesn't incorporate other sensor data
May struggle in feature-poor or dynamic environments
Less suitable for large-scale mapping

Code Comparison

ORB_SLAM2 (C++):

// Feature extraction and matching
void Frame::ExtractORB(int flag, const cv::Mat &im)
{
    (*mpORBextractorLeft)(im,cv::Mat(),mvKeys,mDescriptors);
}

Cartographer (C++):

// Multi-sensor fusion
void LocalTrajectoryBuilder2D::AddRangefinderData(
    const sensor::TimedPointCloudData& rangefinder_data) {
  // Sensor data processing and integration
}

ORB_SLAM2 focuses on visual features for SLAM, while Cartographer is designed for multi-sensor fusion, including lidar and IMU data. ORB_SLAM2 is more specialized for visual SLAM, whereas Cartographer offers a more comprehensive approach to mapping and localization across various sensor types and environments.

hdl_graph_slam

1,974

3D LIDAR-based Graph SLAM

Pros of hdl_graph_slam

Specifically designed for 3D LiDAR data, offering better performance for point cloud processing
Implements loop closure detection using a deep learning-based approach, potentially improving accuracy in complex environments
Provides visualization tools for graph optimization and mapping results

Cons of hdl_graph_slam

Less mature and less widely adopted compared to Cartographer
May require more computational resources due to the deep learning-based loop closure detection
Limited to LiDAR-based SLAM, while Cartographer supports multiple sensor types

Code Comparison

hdl_graph_slam:

void GraphSLAM::add_se3_node(const Eigen::Isometry3d& pose) {
  Vertex* v = graph.add_vertex(pose);
  keyframe_queue.push_back(v);
}

Cartographer:

void PoseGraph3D::AddNode(
    std::shared_ptr<const TrajectoryNode::Data> constant_data,
    const std::vector<std::shared_ptr<const Submap3D>>& insertion_submaps) {
  const NodeId node_id = AppendNode(constant_data, insertion_submaps);
  AddWorkItem([this, node_id]() REQUIRES(mutex_) { OptimizeAfterNode(node_id); });
}

Both projects implement graph-based SLAM algorithms, but hdl_graph_slam focuses on LiDAR data processing, while Cartographer offers a more general-purpose solution for various sensor types and configurations.

Convert designs to code with AI

Introducing Visual Copilot: A new AI model to turn Figma designs to high quality code using your components.

Try Visual Copilot

README

.. Licensed under the Apache License, Version 2.0 (the "License"); you may not use this file except in compliance with the License. You may obtain a copy of the License at

.. http://www.apache.org/licenses/LICENSE-2.0

.. Unless required by applicable law or agreed to in writing, software distributed under the License is distributed on an "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the License for the specific language governing permissions and limitations under the License.

============ Cartographer

Purpose

Cartographer_ is a system that provides real-time simultaneous localization and mapping (SLAM_) in 2D and 3D across multiple platforms and sensor configurations.

|video|

.. _Cartographer: https://github.com/cartographer-project/cartographer .. _SLAM: https://en.wikipedia.org/wiki/Simultaneous_localization_and_mapping

A Note for ROS Users

Cartographer is no longer actively maintained. On rare occasions critical pull requests may be merged, but no new development is currently taking place, including issue response. If you are installing Cartographer in ROS 1 / ROS 2 using a binary package that package is a fork of this repository. The ROS fork of Cartographer is only maintained in a limited capacity. No new development takes place on this fork, but pull requests may be merged at the maintainers' discretion.

The ROS fork of Cartographer, and the ROS wrapper library, can be found at these locations:

Cartographer Fork <https://github.com/ros2/cartographer>_
Cartographer ROS <https://github.com/ros2/cartographer_ros>_

Additional discussion can be found in this ROS Discourse thread<https://discourse.ros.org/t/rolling-and-soon-humble-release-of-both-cartographer-and-cartographer-ros-v2-and-call-for-testing/25137>_.

Getting started

Learn to use Cartographer at our Read the Docs site_.
You can ask a question by creating an issue_.

.. _our Read the Docs site: https://google-cartographer.readthedocs.io .. _creating an issue: https://github.com/cartographer-project/cartographer_ros/issues/new?labels=question

Contributing

You can find information about contributing to Cartographer at our Contribution page_.

.. _our Contribution page: https://github.com/cartographer-project/cartographer/blob/master/CONTRIBUTING.md

Open house slide archive

In the past there had been regular open-for-all meetings to discuss progress and plans for Cartographer. Slides of these Cartographer Open House meetings are listed below.

March 14, 2019: Slides <https://storage.googleapis.com/cartographer-public-data/cartographer-open-house/190314/slides.pdf>_
February 21, 2019: Slides <https://storage.googleapis.com/cartographer-public-data/cartographer-open-house/190221/slides.pdf>_
January 17, 2019: Slides <https://storage.googleapis.com/cartographer-public-data/cartographer-open-house/190117/slides.pdf>_
November 22, 2018: Slides <https://storage.googleapis.com/cartographer-public-data/cartographer-open-house/181122/slides.pdf>_
October 25, 2018: Slides <https://storage.googleapis.com/cartographer-public-data/cartographer-open-house/181025/slides.pdf>_
September 13, 2018: Slides <https://storage.googleapis.com/cartographer-public-data/cartographer-open-house/180913/slides.pdf>_
August 16, 2018: Slides <https://storage.googleapis.com/cartographer-public-data/cartographer-open-house/180816/slides.pdf>_
August 2, 2018: Slides <https://storage.googleapis.com/cartographer-public-data/cartographer-open-house/180802/slides.pdf>_
July 5, 2018: Slides <https://storage.googleapis.com/cartographer-public-data/cartographer-open-house/180705/slides.pdf>_
June 21, 2018: Slides <https://storage.googleapis.com/cartographer-public-data/cartographer-open-house/180621/slides.pdf>_
June 7, 2018: Slides <https://storage.googleapis.com/cartographer-public-data/cartographer-open-house/180607/slides.pdf>_
May 24, 2018: Slides <https://storage.googleapis.com/cartographer-public-data/cartographer-open-house/180524/slides.pdf>_
May 3, 2018: Slides <https://storage.googleapis.com/cartographer-public-data/cartographer-open-house/180503/slides.pdf>_
March 29, 2018: Slides <https://storage.googleapis.com/cartographer-public-data/cartographer-open-house/180329/slides.pdf>_
February 22, 2018: Slides <https://storage.googleapis.com/cartographer-public-data/cartographer-open-house/180222/slides.pdf>_
February 8, 2018: Slides <https://storage.googleapis.com/cartographer-public-data/cartographer-open-house/180208/slides.pdf>_
January 18, 2018: Slides <https://storage.googleapis.com/cartographer-public-data/cartographer-open-house/180125/slides.pdf>_
January 11, 2018: Slides <https://storage.googleapis.com/cartographer-public-data/cartographer-open-house/180111/slides.pdf>_
December 7, 2017: Slides <https://storage.googleapis.com/cartographer-public-data/cartographer-open-house/171207/slides.pdf>_
November 23, 2017: Slides <https://storage.googleapis.com/cartographer-public-data/cartographer-open-house/171123/slides.pdf>_
November 9, 2017: Slides <https://storage.googleapis.com/cartographer-public-data/cartographer-open-house/171109/slides.pdf>_
October 26, 2017: Slides <https://storage.googleapis.com/cartographer-public-data/cartographer-open-house/171026/slides.pdf>_
October 12, 2017: Slides <https://storage.googleapis.com/cartographer-public-data/cartographer-open-house/171012/slides.pdf>_
September 14, 2017: Slides <https://storage.googleapis.com/cartographer-public-data/cartographer-open-house/170914/slides.pdf>_
August 17, 2017: Slides <https://storage.googleapis.com/cartographer-public-data/cartographer-open-house/170817/slides.pdf>_
July 20, 2017: Slides <https://storage.googleapis.com/cartographer-public-data/cartographer-open-house/170720/slides.pdf>_
July 6, 2017: Slides <https://storage.googleapis.com/cartographer-public-data/cartographer-open-house/170706/slides.pdf>_
June 22, 2017: Slides <https://storage.googleapis.com/cartographer-public-data/cartographer-open-house/170622/sildes.pdf>_
June 8, 2017: Slides <https://storage.googleapis.com/cartographer-public-data/cartographer-open-house/170608/slides.pdf>_

Top Related Projects

Convert designs to code with AI

Introducing Visual Copilot: A new AI model to turn Figma designs to high quality code using your components.

Try Visual Copilot