A drone-based image processing method for assessing plant health and deformations in expansive agricultural areas is presented in this work. The YOLOv9 object detection and U-Net segmentation algorithms were used to analyze high-resolution photos taken with a drone system. Although YOLOv9 offered quick and efficient object recognition across large regions, the U-Net segmentation model allowed for in-depth pixel-level analysis, which made it easier to create accurate plant health maps. When combined with geographic information systems (GIS), these maps were an essential tool for precisely pinpointing areas that needed assistance. According to the findings, YOLOv9 consistently identified deformations with good performance measures, such as 82% sensitivity, 79% accuracy, and 78% precision. By accurately and highly accurately differentiating between healthy and harmful areas, U-Net segmentation improved these detections even further. The created technology significantly improves environmental sustainability in addition to increasing resource efficiency and cutting expenses in agricultural operations.