This work is to study earthquake precursors using lineament analysis of five MW3 earthquakes that occurred in 2018 on the territory of the Tashkent geodynamic polygon in Uzbekistan. The Landsat 8 satellite images were processed using the automated lineament detection method in the LEFA software. The processing steps included image preprocessing with graphic filters, edge detection using Canny\'s algorithm, the definition of a vector of linear elements based on the probabilistic Hough transform and combining collinear linear elements into lineaments. The analysis of the cyclicity of precursors before and after earthquakes was based on the study of the distribution of the lineaments trend in the study area using rose diagrams and lineaments density maps. The results showed a change in the dynamics of the lineament structure. The statistical analysis of lineaments revealed that the number and orientation of lineaments changed significantly about 2—4 months before an earthquake, and after 2—3 months, the system gradually returned to its initial state. The rapid increase in lineaments density began almost 20 days before the event, reaching a maximum about 4 days before the earthquake, and its decrease began 16 days after the earthquake. The main trends observed in the lineament map showed dominant trends in NS, EW, and the subdominant directions were in NW—SE, NE—SW, which coincided with main di-rections of the faults.