As a new type of strain measuring instrument, fiber bragg gratings (FBG) strain sensors are gradually used in the field of civil engineering, and the key to their use lies in the deployment of sensors and the calibration of sensitivity coefficients according to different usage requirements and environments. The current conventional FBG installation methods are surface direct bonding and slotting implantation. In order to determine the measurement accuracy of the two installation methods, a calibration method of the unpacked FBG sensitivity coefficient used for the pile sinking test was proposed, the calibration test of the two deployment methods was carried out by using the circular seamless steel tube as the matrix, and the calibration results were compared with the theoretical values. The results show that in the stage of progressive loading, the calibration coefficient obtained by the method of slotted implantation method is smaller than that by surface adhesion method, while the calibration coefficient obtained by the surface adhesion method is closer to the theoretical value. The calibration coefficients obtained by the slotted implantation and surface adhesion methods were 2.84% and 0.33% respectively compared with the theoretical values; In the step-by-step unloading phase, there is a certain lag in the central wavelength of the two deployment methods, but the calibration coefficient is not much different from the loading stage. Therefore, it is necessary to choose a reasonable fiber grating layout scheme and calibration method according to the test environment and within the test tolerance range in the process of using FBG.