In this thesis, we propose a dynamic displacement estimation method for large-scale civil infrastructures based on two-stage Kalman filter and heuristic drift reduction technique. When measuring displacement at a large-scale infrastructures, a displacement sensor have a long measurement distance and limited selections of sensor installation location. RTK-GNSS, therefore, has played a role as the only applicable displacement sensor on civil infrastructures practically due to these limitations. However, RTK-GNSS has a low sampling frequency of 10 Hz, and often suffers from its low stability affected by the number of satellites, location, and surrounding environment. The proposed method combines RTK-GNSS and accelerometer data to estimate the dynamic displacement of the structure with higher precision and accuracy than those of RTK-GNSS, and 100 Hz sample frequency. In the proposed technique, the accuracy of RTK-GNSS displacement measurement is improved by removing the low-frequency noise of RTK-GNSS through the heuristic drift reduction technique proposed in this paper. Then, using the displacement estimated by the heuristic deviation elimination technique, the velocity measured by GNSS, and the acceleration measured with the accelerometer are combined in a two-stage Kalman filter to estimate the dynamic displacement with high accuracy and high sampling frequency. The proposed dynamic displacement estimation technique was verified by lab-scale experiments and field application tests at various long-span bridges such as Youngjong Bridge and Yi Sun-Sin Bridge.