A QGIS plugin for tree monitoring using AI.
MegaDetector-Overhead — The Microsoft open-source AI for overhead wildlife detection. Point-based detection model for aerial and drone imagery, identifying wildlife from above. Maintained by Microsoft AI for Good Lab. Part of the Pytorch-Wildlife ecosystem.
GreenShield Mobile App
This repository provides a tutorial and code for reproducing the data and results presented in the following publication: Automated Mapping of Post-Storm Roof Damage Using Deep Learning and Aerial Imagery: A Case Study in the Caribbean (Kucharczyk, Nesbit, & Hugenholtz, 2025, Remote Sensing).
Semantic segmentation of drone imagery using U-Net deep learning architecture | TensorFlow/Keras implementation with automated preprocessing and multi-class pixel classification | Dron görüntülerinin U-Net ile semantik segmentasyonu
Final Year Project developed for my BSc (Hons) in Computing.
Tiered edge AI land cover classification across ESP32, STM32, and Jetson Nano — the same task, scoped and quantized differently for each board's memory budget.
Drone GeoTIFF üzerinde genç mısır segmentasyonu, sıra analizi ve aday ekim boşluğu tespiti
DNN-derived synthetic vegetation index for agronomic trait prediction from multispectral drone imagery
AI-powered traffic analysis system for detecting and tracking road users from aerial drone footage.
Deep learning pipeline (YOLOv5 + Detectron2) for detecting maize tassels in UAV imagery — CSE499 senior design project.
A modern desktop app for calculating areas from UAV-captured GeoTIFFs and shapefiles. Features support for alpha masking, automatic UTM projection, clean map overlays, and precise dimension reporting.
Genomic and phenomic prediction of agronomic traits and malt quality in NY Winter Malting Barley
Technical and architecture documentation of a precision-agriculture object detection pipeline (YOLO + SAHI, geospatial export, COCO evaluation). Sanitized public release — documentation only, no source code or client data.
End-to-end drone remote sensing platform: schema-agnostic data pipeline, GeoLibre-based GIS mapping with IDW heat-surface interpolation, OpenCV computer vision for land-cover classification, and AI-assisted automated reporting — built with Streamlit.
End-to-end MLOps pipeline for 5-class semantic segmentation of drone imagery using nnU-Net (PyTorch). Features DVC, Docker, BentoML serving & CI/CD.
UNet-based density estimation framework for automated elephant counting using UAV imagery.
Civilian computer-vision research pipeline for object detection, embedding-based target matching, and lock-state management in recorded drone imagery.
Drone-based livestock detection system using YOLOv8 for identifying and counting sheep in aerial imagery. Includes dataset details, training pipeline, model weights, and Colab notebook.
DroneSeg is a full-stack semantic segmentation platform for drone/aerial imagery. Upload drone photos, run SegFormer-B2 deep learning inference to generate land-cover classification masks with bounding boxes and visualize results on an interactive map and GeoJSON export.
To associate your repository with the drone-imagery topic, visit your repo's landing page and select "manage topics."