This package subscribes twist stamped message from two different IMUs, removes noise using a kalman filter, fuses using a complimentary filter then calculates the odometry (x,y and theta). It then publishes the calculated Odometry and fused velocities. The code is implemented in C++ and works with ROS2.
Next step is to replace the complimentary filter with kalman filter for fusion and implement the same for all velocities (x,y,z,wx,wy,wz) to track in 3D and generalize for N number of sensor inputs.
Traced odometry path using two sources of velocities:
