One of the most vital parts of an aero-engine is the rotor, and the working state of a rotor blade is crucial for determining the performance and safety of the engine. Blade tip timing (BTT), as an efficient and non-invasive blade vibration measurement technology, has realized the online monitoring of blade vibration. Online monitoring requires rapid and accurate extraction of blade vibration parameters. By identifying the resonance of blades, vibration parameters can be effectively extracted, such as amplitude, engine order (EO), natural frequency and other information. Through finite element simulation, it is found that when the blade is in synchronous resonance, the vibration phase measured by two probes will be different, as long as the angle between two probes has no integer multiple relation with 360°/EO. Based on this, a phase-offset resonance identification (PHRID) method was presented to identify synchronous resonance of blades. This method identifies resonance by extracting the vibration trend of probes with a special installation angle. Finally, rotor blades were tested in the experiment and compared with the strain data, and the frequency identification error within 1%. The core of this method is the trend extraction of the BTT signal of a certain probe, which greatly reduces the calculation amount and is promising to realize online monitoring. And in the case of extremely limited installation, only two probes with special angles are needed to achieve accurate extraction of resonance frequency.