Overview

Architecture, API, and implementation notes for my public C++/Eigen IMU+GPS error-state Kalman filter project.

Conventions

Conventions discusses frames, notation, quaternion representation, and sign conventions.

Inputs

Inputs discusses filenames, runtime selection, time properties, field schemas, field meanings, and dataset-specific edge cases.

Synthetic GPS

Synthetic GPS discusses how the pose stream is turned into the synthetic horizontal GPS measurements used by startup and update.

Initialization

Initialization discusses how the app turns early IMU, synthetic GPS, and wheel-speed samples into the startup handoff state for the live filter.

ESKF

ESKF discusses the filter state, IMU propagation, 2D GPS update, and the post-startup runtime loop.

Outputs

Outputs discusses the written run artifacts and the successful-run summary.

API

API discusses the public C++ declarations by Groups and Types.