Abstract: Mastering the characteristics and process of landslide surface deformation is an important prerequisite for revealing the formation mechanism of a landslide and implementing effective prevention and control. With the rapid development of UAV photogrammetry technology, it has become possible to achieve two-dimensional deformation monitoring and analysis of a single landslide surface using high-resolution UAV time-series digital orthophotos(DOMs)combined with pixel offset tracking(POT)analysis; however, a complete technical method system has not yet been formed. In this paper, the method of UAV combined with POT technology was used to monitor the Tanjiawan landslide in the Three Gorges Reservoir area over three and a half years. The applicable conditions of UAV on-site monitoring were determined, and the optimal parameters for two-dimensional deformation field analysis were obtained based on COSI-Corr software. The accuracy of the method was verified using GNSS monitoring data. A set of technical methods for monitoring and analyzing the surface deformation of a single long-term creeping landslide under vegetation cover was established. Finally, the surface deformation characteristics of the Tanjiawan landslide were analyzed. The results showed that the UAV combined with POT technology could effectively monitor the key displacement area and direction of the landslide with high accuracy. However, due to the influence of vegetation cover, the technology was limited to the identification of displacement deformation below 20 mm, and the DOM acquisition time interval was less than one year. In the COSI-Corr software, the initial and final windows were set to 32×32, the displacement step was set to 6, and the DOM selected the band and used a pixel value of 0.12-0.2 m to obtain the best surface displacement deformation monitoring results. The deformation of the Tanjiawan landslide was mainly concentrated in sliding bodies No. Ⅰ, Ⅱ, and Ⅲ, and the deformation had regional specificity.