Expand description
Visual odometry and SLAM building blocks for kornia-rs.
Re-exports§
pub use frame::Frame;pub use loop_closure::LoopClosingConfig;pub use loop_closure::LoopClosureEvent;pub use sensor_rig::ImuCalibration;pub use sensor_rig::SensorRig;pub use system::SlamConfig;pub use system::SlamSystem;pub use system::TrackingResult;pub use system::TrackingStatus;pub use tracking::KeyframePolicy;pub use tracking::MapProjectionEstimator;
Modules§
- frame
- Core frame-domain types shared across the crate.
- initialization
- Visual and inertial initialization for the SLAM system.
- loop_
closure - Loop closing: place recognition, geometric verification, consistency, fusion, and pose-graph optimization.
- mapping
- Mapping: the map itself, plus the algorithms and worker that operate on it.
- sensor_
rig - Fixed sensor calibration, independent of the estimated SLAM state.
- stereo
- Stereo correspondence: per-keypoint disparity and metric depth from a rectified, row-aligned stereo pair.
- system
- SLAM runtime: orchestrates tracking, mapping, and state transitions.
- tracking
- Frame-to-map tracking: pose estimation, optical flow, the motion model and the keyframe and loss-recovery policies.
Structs§
- OrbFeatures
- ORB features extracted from a single frame.