METHODOLOGY FOR THE AUTOMATED DETECTION OF ANOMALOUS GEOSPATIAL ZONES IN SATELLITE IMAGERY USING STATISTICAL ANALYSIS AND A CUSTOM QGIS PLUGIN

Background. This article presents a methodology for the automated detection of anomalous geospatial zones, implemented as a plugin for the QGIS geographic information system. The developed tool enhances the efficiency of spatial analysis and enables the rapid identification of areas with potential changes for monitoring natural and anthropogenic processes. Methods. The proposed approach is based on thresholding and statistical analysis of satellite imagery within the QGIS environment. The plugin provides interactive adjustment of image processing parameters and automatically detects geodynamic anomalies, which are then vectorized and delivered to the user for further analysis. The algorithm utilizes Python libraries (NumPy, SciPy, GDAL, PyQt, QGIS API) to handle various types of satellite data and applies standard deviation-based criteria to identify anomalous areas. Results. The testing of the plugin developed by the authors confirmed its effectiveness in processing satellite imagery types such as InSAR, thermal infrared (TIR), and NDWI-based images. The plugin successfully identified areas of vertical displacement of the Earth's surface, detected thermal anomalies, and delineated regions with moisture deficits. This approach substantially improves the accuracy of geospatial analysis. Conclusions. The developed plugin is an effective tool for the automated monitoring of changes in the Earth's surface and the assessment of hydrogeological conditions. Its integration within the QGIS environment enables the efficient adjustment of analysis parameters and the generation of results in vector data format. Plugin testing confirmed its practical value and revealed potential directions for further improvement, particularly regarding the separate processing of positive and negative displacement values to enhance the accuracy of anomaly interpretation.