To reduce the output signal noise of Hemispherical Resonator Gyro (HRG) in dynamic environment and improve the precision of inertial navigation system,a filtering algorithm based on Lifting Wavelet Transform (LWT) is proposed.The signal is decomposed with wavelet transform lifting algorithm,and then the low-frequency wavelet coefficient is processed with Forward Linear Prediction (FLP) filter,to further remove the low-frequency interference and extract the desired signal.The high-frequency wavelet coefficient is directly to be zeroed or processed according to certain threshold rules,which significantly improves the accuracy of reconstructed signal,and realizes removal of the gyro noise.A comparison of the new algorithm,Kalman filtering algorithm and FLP filtering algorithm is made in detail,as well as an analysis of their de-noising effect.Its verified by the simulation results and actual test data that the new algorithm can effectively reduce the output noise of HRG in dynamic environment,the standard deviation is reduced by about 50% and the operation time is shortened by 30% as compared with Kalman filtering.This filtering algorithm is not only of fast speed and high accuracy,but also with a small amount of calculation and good anti-interference performance,enabling to effectively track the desired signal after filtering.