Aiming at the multi-functional properties of Pb-based lanthanumdoped zirconate titanates (PZT) sensors in concrete for structural health monitoring, which enable the sensors to simultaneously receive signals with different functions, a method for extracting the signals for variant purposes in the multi-functional PZT sensors is proposed. Because there is a difference in frequency ranges from different function signals, the vibration signal related to the overall structure performance and the acoustic emission signal associated with local damages are acquired based on the Mallat algorithm. The correctness of the extracted signals is verified by comparing with those from accelerometers and acoustic emission sensors. In addition, the proposed method is applied to the seismic damage experiment of a reinforced concrete frame-shear structure. The experimental results show that the vibration signal acquired by the proposed method can abstract the frequency of the structure. Meanwhile, the acoustic emission signal abstracted from PZT sensors can monitor the released energy caused by local damage. It can be concluded that the vibration signals and acoustic emission signals can be extracted using the proposed method, and the evaluation of the overall dynamic performance and local damage can be realized.