Course Hive
Search

Welcome

Sign in or create your account

Continue with Google
or
Remote Sensing Image Segmentation with Meta's SAM 3
Play lesson

GeoAI Tutorials - Remote Sensing Image Segmentation with Meta's SAM 3

4.0 (2)
23 learners

What you'll learn

This course includes

  • 20 hours of video
  • Certificate of completion
  • Access on mobile and TV

Summary

Keywords

Full Transcript

Unlock powerful GeoAI workflows with SAM 3! In this step-by-step tutorial, I demonstrate how to segment remote sensing imagery using text prompts and bounding boxes, powered by Meta’s SAM 3 (Segment Anything Model 3). You’ll learn how to run image segmentation on satellite and aerial imagery, extract objects of interest, and export the results to geospatial formats like GeoTIFF for further GIS or Python analysis. Whether you work in remote sensing, geospatial AI, earth observation, or environmental mapping, this video will walk you through a practical, reproducible workflow using open-source tools from the SamGeo (segment-geospatial) library. 🔗 GitHub Repository (SamGeo): https://github.com/opengeos/segment-geospatial 🔗 Notebook Example: https://samgeo.gishub.org/examples/sam3_image_segmentation 🔗 Meta SAM 3 Overview: https://ai.meta.com/sam3 00:00 Introduction: Segmenting remote sensing imagery with SamGeo 00:08 What is SAM 3 & how it compares to SAM 1 & 2 01:24 Getting started: Tools & website overview 01:49 Running SAM3 in Google Colab or locally 02:05 Requesting model access from Hugging Face 03:05 Installing SamGeo3 and dependencies 04:04 Using sample imagery and rendering it on map 05:00 Hugging Face login & token setup 05:58 Creating SAM model instance (backend & device options) 07:14 GPU requirements and model size 08:54 Loading the model and imagery into GPU 09:33 Text prompt segmentation (e.g. "buildings") 10:55 Visualizing segmented objects with bounding boxes 12:23 Customizing mask visualization 13:50 Using bounding boxes instead of text prompts 15:01 Drawing bounding boxes and getting coordinates 16:23 Generating mask from bounding boxes 17:51 Comparing segmentation from prompt vs box 18:56 Saving masks and segmentation results 20:15 Visualizing confidence scores and segmentation overlay 22:27 Filtering segmentation with min object size 24:27 Final visualization and comparison 24:59 Reference to Meta’s official SAM-3 GitHub repo 25:36 Outro: What’s next (non-geospatial data & interactive segmentation) 📺 GeoAI Playlist: https://www.youtube.com/playlist?list=PLAxJ4-o7ZoPcvENqwaPa_QwbbkZ5sctZE 📚 Get my new Books: 📘 Introduction to GIS Programming: A Practical Python Guide to Open Source Geospatial Tools 👉 Amazon: https://amazon.com/dp/B0FFW34LL3 👉 Leanpub: https://leanpub.com/gispro 📘 Spatial Data Management with DuckDB: From SQL Basics to Advanced Geospatial Analytics 👉 Amazon: https://amazon.com/dp/B0G2JFMFFC 👉 Leanpub: https://leanpub.com/duckdb 👋 Let’s Connect: YouTube: https://youtube.com/@giswqs LinkedIn: https://www.linkedin.com/in/giswqs Twitter: https://twitter.com/giswqs Facebook: https://www.facebook.com/groups/opengeos

Course Hive

Continue this lesson in the app

Install CourseHive on Android or iOS to keep learning while you move.

Related Courses

FAQs

Course Hive
Download CourseHive
Keep learning anywhere