JKalman is a Kalman Filter library written by Petr Chmelar.
For the original source code see: https://sourceforge.net/projects/jkalman/
Copy this directory into processing/libraries
// 1-dimensional example
JKalman kalman = new JKalman(2, 1);
// Transition matrix (model dynamics)
Matrix transition = new Matrix(new double[][]{{1, 1}, {0, 1}});
kalman.setTransition_matrix(transition);
// Initial state estimate (can be zero or some initial guess)
Matrix initial_state = new Matrix(new double[][]{{0}, {0}});
kalman.setState_pre(initial_state);
// Example measurement loop
Matrix measurement = new Matrix(1, 1); // One measurement
double[] noisyMeasurements = new double[]{2, 3, 2.5, 3.2, 2.8}; // Example measurements
for (double measurementVal : noisyMeasurements) {
measurement.set(0, 0, measurementVal); // Set the current measurement value
// Correct the state estimate with the new measurement
Matrix corrected_state = kalman.Correct(measurement);
System.out.println("Corrected State: " + corrected_state.get(0, 0));
// Predict the next state
Matrix predicted_state = kalman.Predict();
System.out.println("Predicted State: " + predicted_state.get(0, 0));
}For 2-dimensional (x,y) example see: examples/Test.pde
gradle -qgradle javadocGNU General Public License version 2.0 (GPLv2), Intel Open Source License