Object Detection & Coordinate Extraction from Aerial Imagery

Описание: In this video, I demonstrate how objects such as cars can be detected directly from aerial and satellite images and then converted into real-world map coordinates. Instead of stopping at visual bounding boxes, the detected objects are transformed into accurate geographic shapes and exported as GeoJSON, making them ready to use in maps and GIS applications. You’ll see the full workflow, starting from aerial imagery, moving through object detection, and ending with an interactive map that shows each detected object in its correct location on the Earth. This approach is useful for satellite image analysis, smart city projects, traffic monitoring, urban planning, and AI-powered geospatial workflows. The video keeps things practical and results-focused, making it easy to understand how aerial images can be turned into meaningful, location-aware map data.

Colab Notebook: https://colab.research.google.com/dri...

Roboflow Universe: https://universe.roboflow.com/chargep...

Test GeoTIFF Image: https://drive.google.com/file/d/1fQ-c...

Follow me on:
LinkedIn:   / rahim-biswas  
GitHub: https://github.com/Rahim-Biswas


00:00 Introduction
04:05 Model Training
07:21 Model Testing
11:23 Coordinate Extraction
12:49 Plotting on Map

‪@spatia-ai‬

#gis #deeplearning #remotesensing #geoai #geospatial #spatial #ai #map #objectdetection #cardetection #maps