A METHOD FOR MOTION ESTIMATION IN A VEHICLE, CORRESPONDING DEVICE AND COMPUTER PROGRAM PRODUCT

A method, comprising: receiving motion data from at least one inertial sensing unit (11) and absolute positioning data from a global navigation system (13) in a wheeled vehicle; applying pre-processing (21, 22, 23) to motion data as a function of absolute positioning data, obtaining pre-processed motion data and at least one estimated velocity value (v, v) of the wheeled vehicle; applying pipeline data processing (24) to pre-processed motion data as a function of the at least one estimated velocity value (v, v) of the wheeled vehicle, obtaining a calculated virtual vehicle velocity (v) as a result; applying position-velocity-attitude, PVA, filtering (23) to pre-processed motion data and to at least one of absolute positioning data and the calculated virtual vehicle velocity (v), obtaining fused data indicative of position, velocity and attitude information of the wheeled vehicle as a result. The pipeline data processing (24) comprises: high-pass or low-pass filtering (25) pre-processed motion data and obtaining filtered data as a result; applying frequency-domain transformation (26) to filtered data and obtaining, as a result, frequency analyzed data comprising individual spectral power density signals including a set of power-maximizing frequencies; calculating (27, 28), as a function of said set of power-maximizing frequencies, a wheel angular frequency signal (f) and a wheel size signal (C); calculating (29) said virtual vehicle velocity (v) as a function of said wheel angular frequency signal (f) and said wheel size signal (C).