Abstract: The optimal estimation of the mass unbalance of spherical rotor is got by adopting nonlinear Kalman filter to process time series data of hunting trace of the spherical rotor. The nonlinear time series data are obtained by using two dimensional monochrome CCD and PCI image board which time-sequentially capture the hunting rotor’s image in which a certain identification point characterized by its grey level is recorded of its position on the rotor by using image processing technique. The system error and random error are analyzed and the experimental results which prove the system has the merits of self-operated measuring, timesaving run duration, high precision are given.

Key words: CCD, Image processing, Kalman filter, Mass unbalance measurement, Spherical rotor