cartographer-project/cartographer
Cartographer
Cartographer is a cross-platform robotics library and framework for simultaneous localization and mapping in 2D and 3D spaces. It functions as a real-time mapping engine that constructs environmental maps while tracking a device's position and orientation using continuous sensor data processing.
The system implements real-time SLAM to generate precise maps for autonomous navigation. It utilizes a localization system that determines a device's state within a mapped environment across different hardware platforms and sensor configurations.
The framework covers spatial estimation through non-linear optimization, loop closure, and multi-sensor data fusion. It manages environmental representations using probability-based grid mapping and a hierarchical submap-based approach to handle large-scale mapping.
The library includes a build configuration system to facilitate integration into external software projects.
Features
- Real-Time SLAM - Implements a real-time mapping engine that constructs environment maps while simultaneously tracking device location.
- Loop Closure Detection - Provides mechanisms to detect previously visited locations to correct accumulated spatial drift.
- Localization Systems - Provides a localization system to determine the position and orientation of a device within a mapped environment.
- Sensor Fusion - Combines asynchronous data from multiple hardware sensors into a single consistent state estimate for localization.
- SLAM Algorithms - Implements a comprehensive framework for simultaneous localization and mapping in 2D and 3D spaces.
- Navigation Frameworks - Generates precise environment maps required for autonomous robot navigation without human intervention.
- Mapping Engines - Ships a real-time mapping engine that constructs environmental maps through continuous sensor data processing.
- Spatial Mapping - Creates precise 2D and 3D digital spatial representations of physical areas using processed sensor data.
- Pose Optimization - Implements non-linear least squares optimization via Ceres to refine device trajectories and environmental maps.
- Robotic Platforms - Provides a cross-platform library for integrating mapping and localization across various hardware platforms and sensor setups.
- Probabilistic Occupancy Grids - Represents the environment as a grid of occupancy probabilities to handle sensor noise and measurement uncertainty.
- Scan Matching - Aligns incoming sensor data with local submaps through iterative optimization to determine device position.
- Submap-Based Mapping - Organizes the environment into small local submaps to maintain computational efficiency during large-scale mapping.
- SLAM and Odometry - Real-time 2D and 3D SLAM for multiple sensor configurations.
- SLAM and Mapping - Real-time 2D and 3D SLAM across multiple platforms.
