Abstract: Interferometry Synthetic Aperture Radar (InSAR) technology has become an important tool for monitoring surface deformation with its all-day, all-weather ground monitoring and high spatial resolution, and has been widely applied to seismic deformation monitoring. Currently, the most commonly used technique for coseismic deformation monitoring is differential InSAR (D-InSAR). However, the traditional D-InSAR is susceptible to spatial and temporal uncorrelation in areas such as waters and densely vegetated areas, resulting in serious contamination of the coseismic deformation field. In addition, the seismic deformation field sometimes contains obvious atmospheric delay that can affect source parameter inversions. Therefore, improving the quality of coseismic deformation is of great significance for future seismic deformation monitoring and parameter inversion. The multi-temporal InSAR (MT-InSAR) technique, which is widely used in inter-seismic and post-seismic deformation monitoring, is able to suppress the effects of spatiotemporal decorrelation and atmospheric noise. In this paper, we propose a high-precision coseismic deformation monitoring method based on time-series InSAR analysis to obtain high-precision coseismic deformation results. The accuracy of coseismic deformation field is mainly improved by selecting appropriate interferograms and selecting stable points. With the support of sufficient Sentinel-1A/B satellite SAR data, numerous interferograms were generated using a large number of pre- and post-earthquake images. Interferograms that are less affected by errors are selected for study according to certain criteria to reduce the impact caused by atmospheric delay errors. At the same time, setting threshold to select stable point target to improve the accuracy of deformation field. Taking the 2018 Hualian M_W6.4 earthquake in Taiwan China as an example, the data processing flow of high-precision coseismic deformation monitoring method is introduced in detail. Compared with the results of the traditional D-InSAR method, the proposed method can reduce the noise error and improve the signal-to-noise ratio of coseismic deformation. The high-precision coseismic deformation monitoring method is applied to obtain the seismic deformation of 14 different magnitudes and locations. The deformation results show that the method can improve the accuracy of deformation field by selecting stable points, and it is generally applicable.