Stochastic resonance (SR) in the bistable system can reduce the noise level, reveal the waveform’s outline prominently, and improve the quality of the system output. Under the support of SR techniques, the extraction accuracy of the civil aero-engine rotor vibration information is lifted, especially when treating the signals contaminated by heavy noise. Results indicate that the SR signal preprocessing method via tuning the relative parameters is effective to reflecting the real vibration of the aero-engine rotor. It can also monitor the short-term failure phenomenon in the sampling data.