The signal phase differences of Coriolis sensor and the mass flow are proportional. To improve the measurement accuracy of flow signal processing for Coriolis mass flowmeter, a novel method based on Hilbert Transform algorithm was proposed. The main method is as follows, two signals enhanced by LZ-filter with noise-canceling were considered as Hilbert Transform, then, the phase difference of the two filtered signals was calculated by a triangle characteristic of sin function. Synchronously, the instantaneous frequency was estimated by the structure of the analytical signal. Finally, the mass flow value was obtained by calculating the phase difference. The simulation experiments and Digital Signal Processing system (DSP) verification testing demonstrate that the new signal processing method, which is the LZ algorithm, has better characteristics of real-time and high-precision than the others. The experimental results show that the accuracy of phase difference measurement is 0.02% and the tracking accuracy of frequency is better than 0.07%.