<p>Pipeline structures are crucial in the transportation of national strategic and public resources such as oil, gas, and water. Due to the erosion caused by the transported media and the surrounding environment, pipe structures are prone to damage such as micro-cracks, frictional wear or corrosion-induced perforations, which may ultimately lead to the leakage of the internal media. It is thus particularly important to establish a smart health monitoring system to efficiently and promptly identify damages in pipeline structures. Therefore, this paper proposes a semi-automatic structural damage identification method based on monitoring data from fiber Bragg grating (FBG) sensors and modal parameters identified by the improved covariance-driven stochastic subspace identification (SSI-COV) method. Two types of packaged FBG sensors have been adopted to measure data of the pipe under natural excitation. Results demonstrate that the use of empirical formulas for system order and delay time in the SSI-COV method, combined with stability diagrams for eliminating spurious modes, enables the filtered data to discard noise-dominated information, thereby improving both computational efficiency and damage identification accuracy. When the natural excitation is insufficient, the identified modal parameters may contain significant errors. Therefore, an optimization method is further introduced to improve the accuracy of the identified modal parameters. Experimental results demonstrate that the damage index (<i>DI</i>) calculated by using the modified modal parameters can effectively locate damage at the peak value. The structural damage identification analysis confirms that the proposed method can efficiently identify damage in a short time with strong robustness. With short data processing time and high efficiency, this method supports the development of real-time health monitoring systems and enables fast damage detection in pipeline structures.</p>