Implementation of localization using sensor fusion of GPS/INS/compass through an error-state Kalman filter.
The MATLAB code borrows heavily from Paul D. Groves' book, Principles of GNSS, Inertial, and Multisensor Integrated Navigation Systems, his MATLAB code is marked as his, and is held under the BSD license.
This code is very much a work-in-progress as I transition from a MATLAB implementation to C++. Until the code is in a stable state, I will make changes based on my own needs, with complete disregard for backwards compatibility. If you liked it, then you shoulda used a fork on github. Following the initial development, I may make it more portable.
PLEASE NOTE: I cannot provide my testing data as it was collected using a vehicle owned by General Motors. Please, refer to the comments in the initialization scripts for a description of the data format and you should be able to adapt your own data. Alternatively, the book I used comes with a CD-ROM that has his code (fully working filter) along with a system for generating test data.