07-08, 13:30–17:30 (US/Pacific), Room 318
Drone imagery is more widely available than ever before, allowing the public to capture ultra high-resolution Earth images with hobbyist drones. In this workshop, we will explore drone imagery with Python tools such as geopandas, OpenCV, rasterio, numpy, and shapely. Afterwards, we will assess urban green spaces, focusing on counting trees and estimating their role in capturing carbon to fight climate change. This practical exercise will not only enhance our understanding of urban ecology, but also highlight the importance of trees in urban planning and environmental sustainability.
The tutorial will be a combination of slides and hands-on live coding with real-world data in Jupyter Notebooks and will cover the following topics:
-
Introduction to Geospatial Imagery and Basics Image Processing (First 1/3 of the Tutorial)
- Briefly introduce and explain the structure of the workshop and give motivation for drone imagery in the fight against climate change.
- Inform attendees about accessing Jupyter notebooks hosted on Google Colab, including all hands-on material for the workshop, drone imagery, metadata, and additional materials.
- Start with an introduction to drone technology and its application in environmental analysis, focusing on urban green spaces.
- Cover basic concepts of working with drone imagery, including understanding drone capabilities, image resolution, and data types.
- Introduce the Python libraries that will be used throughout the workshop, such as rasterio for raster data manipulation, numpy for numerical analysis, openCV for computer vision tasks, shapely for geometric operations, geopandas for geospatial data processing, and computer vision libraries for object detection and identification. Explain the significance of each in the context of drone imagery analysis. -
Hands-On Drone Data and Machine Learning Exercise (Remaining 2/3 of the Workshop)
- Transition into a more interactive session, where participants will follow along with pre-made Jupyter Notebooks hosted on Google Colab. This session is designed to be accessible and interesting for all skill levels, from beginners to experienced Python users. Here we will be focusing on tools we will use in the final hands-on exercise, counting trees and estimating carbon capture.
- Introduction to reading geospatial data, extracting images, and composing a scene.
- Dive deeper into geospatial analysis and computer vision packages, showcasing how to apply these tools to drone imagery.
- Explore basic computer vision applications, focusing on image processing techniques and object detection using models to identify and count trees in urban settings.
- Discuss how to interpret the results and estimate a region’s carbon capture.
- Hands-on exercise: Here we will be combining everything we learned in the workshop thus far to create a tool to estimate a city’s ability to mitigate climate change from its greenery.
Some knowledge of image processing techniques and geospatial data is welcomed, but not required.
Installation Instructions –None. Everything is hosted on Google Colab!
Kevin Lacaille works as a senior software engineer at Spexi Geospatial, a crowd-sourced drone imagery company, where he combines the worlds of GIS and computer vision to create an open marketplace for ultra high-resolution Earth imagery. Kevin specializes in building image processing solutions for science and educational teams and communicates his workflows with blog posts and webinars. Kevin has over 8 years of public speaking experience ranging from conferences to academic lecturing. Recently, he presented hands-on tutorials at PyCon 2023 and SciPy 2022. In the past, Kevin has presented at L3Harris Engineers Week 2021, and the Canadian Astronomical Society Conference 2019. Portfolio: www.lacaille.dev