The application of improved SURF optical flow algorithm in seismic support safety monitoring system

Frequent earthquakes pose a major threat to the safety of building structures. As a critical component of earthquake-resistant systems, dynamic health monitoring of seismic isolation bearings is essential. However, existing monitoring methods are limited by high cost and poor environmental adaptability, and it is difficult to meet real-time requirements under complex conditions. Therefore, this study combines the stable acceleration feature point matching with the optical flow algorithm, and uses the proportional purification method and the bidirectional purification method to effectively match and purify the feature points. By optimizing the feature point matching and dynamic tracking accuracy, a seismic bearing safety monitoring system based on the improved feature optical flow algorithm is designed. The experimental results show that the improved model achieves a feature recognition rate of >90 %, and the number of feature tracking losses is <70. At the same time, the displacement time history curve of the seismic isolation bearing and the laser displacement sensor has the highest degree of fit, the error of the horizontal displacement peak is 2.13 %, and the absolute error is <0 .3mm. Compared with traditional methods, this system achieves high-precision real-time monitoring at a low cost, providing reliable technical support for the health management of seismic isolation bearings, and also opens up a new direction in the field of structural health monitoring.