AGRICULTURAL LAND AREA CALCULATION METHOD BASED ON GROUND SAMPLE DISTANCE (GSD) VALUES IN SEGMENTED GEOSPATIAL IMAGES

Abstract

The present invention proposes an automated computational method for precise calculation of agricultural land area using real geospatial parameter extraction. The method flow includes a data collection step (100) for conducting drone flights and acquiring aerial images of agricultural land; a data preprocessing step (200) for image stitching and AI segmentation to form a binary mask of the land; a computational processing step (300) that extracts pixel resolution values ​​from metadata to calculate the Ground Sample Distance (GSD) without rounding and performs connected component analysis for identification of agricultural land objects and automatic noise filtering of objects below 1.0 ð 'š2; and a visualization step (400) that integrates the results of the land area calculation into a database and a WebGIS interface.This integration between real metadata extraction and pixel scanning algorithms enables automatic, fast, and accurate calculation of the area of ​​each land plot, while eliminating the margin of error of conventional manual digitization methods and the interference of non-land pixel noise.