Abstract: Due to unavoidable factors such as line-of-sight obstruction, complex working conditions, and narrow spaces, it is often difficult to accurately perceive the vibration state of equipment under completely obstructed conditions such as behind cover or enclosed spaces. This paper proposes a passive wireless vibration isolation detection method based on radio frequency tag to meet the high-precision detection requirements of equipment vibration signals under complete occlusion conditions. Leveraging the lightweight and miniaturized characteristics of passive radio frequency tags, the method adapts to detection conditions in complex and confined spaces without interfering with the state of the measured target. In response to complex complete occlusion conditions, this paper establishes a typical partition-wall detection scenario and develops a high-precision vibration perception model based on tag phase information. This model establishes a direct association between the phase information of the tag and the vibration signal. By placing the phase information that carries vibration characteristics in the high-power path and simplifying the signal propagation route, the model enhances detection sensitivity and mitigates the influence of multipath effects on detection accuracy. Experimental results demonstrate that the proposed method yields measurements highly consistent with those obtained from a laser displacement sensor under line-of-sight conditions, while simultaneously enabling isolation-based vibration detection. In addition, to account for various constraints present in real-world operating environments, this study designs several representative fully occluded scenarios, including cases with directional misalignment between the antenna and the vibration measurement point, as well as configurations where the antenna and the measurement point are situated in separate enclosed spaces. The results show that this method can ensure stable high-precision detection performance even under the complete obstruction of multiple walls in the complex working conditions mentioned above. The research results offer a new high-precision detection scheme for equipment vibration in non-line-of-sight scenarios, particularly in completely occluded and enclosed spaces.