Orientation Filter

This project implements an orientation filter for a MARG (Magnetometer, Accelerometer, and Gyroscope) sensor array in Swift. The filter compensates for magnetic distortion and gyroscope drift, ensuring accurate orientation estimation.

Features

Gyroscope Drift Compensation: Reduces errors caused by gyroscope drift.
Magnetic Distortion Compensation: Adjusts for magnetic interference to provide accurate orientation data.
Quaternion-Based Orientation Estimation: Uses quaternions for smooth and continuous orientation representation.

Requirements

Swift 2.0 or later

Installation

Include the Filter class in your Swift project.

Usage

To use the orientation filter, create an instance of the Filter class and update it with sensor readings.

Example

import Foundation

// Create an instance of the Filter class
let filter = Filter()

// Update the filter with current sensor readings
filter.filterUpdate(
    w_x: gyroX, w_y: gyroY, w_z: gyroZ, 
    a_x: accelX, a_y: accelY, a_z: accelZ, 
    m_x: magX, m_y: magY, m_z: magZ
)

Parameters

w_x, w_y, w_z: Gyroscope readings (in radians per second).
a_x, a_y, a_z: Accelerometer readings.
m_x, m_y, m_z: Magnetometer readings.

How It Works

Normalization: The accelerometer and magnetometer readings are normalized.
Objective Function and Jacobian: Computes the objective function and its Jacobian to estimate the direction of the gyroscope error.
Gradient Descent: The filter uses gradient descent to minimize the error in the orientation estimate.
Quaternion Update: Updates the quaternion representing the current orientation.
Gyroscope Bias Correction: Corrects for any bias in the gyroscope readings.

Contributing

Contributions are welcome. Please submit a pull request or open an issue to discuss any changes.

Author

Adam Leitgeb
Petr Chmelař

License

This project is licensed under the MIT License. See the LICENSE file for details.

foshlabs/OrientationFilter