Archinect - News

Google updates Open Buildings data sets using AI

2024-10-28T10:46:00-04:00

In 2021, the Google Research Africa team launched Open Buildings, an open-source dataset of building footprints across the Global South produced using AI and high-resolution satellite imagery. The team had a simple vision: to fill a major gap in data for population and density in the developing world. Now in its third version, their dataset contains polygons for 1.8 billion buildings over an area of 58 million km² in Africa, South and Southeast Asia, Latin America and the Caribbean.

The data is useful in determining population size and other factors to solve urban density problems.

Google's Research product manager Olivia Graham says: "About 2.5 billion more people could move to cities by 2050, most of them in the Global South — this could be a real step change for governments and organizations working through that growth. If a city is planning where to put essential services like healthcare and education, or where to develop infrastructure like water and energy supplies, this dataset shows the areas that are actively growing."

The data can be explored via a searchable map using that's accessed here.

Artificial Intelligence helps mapping urban trees (all of them)

2019-01-07T14:19:00-05:00

To train the model, he identified known locations of tree canopy using lidar data and NAIP imagery over California. Using that as ground truth, the model was trained to classify which pixels contain trees in the corresponding satellite images. The result is a machine-learning model that has learned to identify trees just using four-band high-resolution (~1 meter) satellite or aerial imagery—no lidar required!

Former New York Times cartographer Tim Wallace describes how his current firm, Santa Fe-based Descartes Labs, has built a machine learning model to identify tree canopy from satellite imagery thus making accurate mapping of trees and urban forests far more accessible to cities worldwide.

San Francisco Open Forest Map tree inventory (point data) in comparison with the Descartes Labs tree canopy layer (image data). Image: Tim Wallace/Descartes Labs

"The ability to map tree canopy at a such a high resolution in areas that can’t be easily reached on foot would be helpful for utility companies to pinpoint encroachment issues—or for municipalities to find possible trouble spots beyond their official tree census (if they even have one)," writes Wallace. "But by zooming out to a city level, patterns in the tree canopy show off urban greenspace quirks. For example, unexpected tree deserts can be identified and neighborhoods that would most benefit from a surge of saplings revealed."

New York...

Google Maps is integrating buildings and architectural details, far surpassing other map makers such as Apple & Bing

2017-12-21T16:51:00-05:00

Justin O'Beirne lays out years worth of research on mapping technologies in his essay Google Map's Moat. O'Beirne reveals,"Over the past year, we’ve been comparing Google Maps and Apple Maps [...] The biggest difference is the building footprints: Google seems to have them all, while Apple doesn’t have any. [...] The buildings are a new thing, and I’ve been watching Google gradually add them over the past year."

With plenty of map screenshot GIFs, O'Beirne illustrates how Google has integrated architecture (including sheds, trailers, garages, etc.) with fairly accurate architectural detail just in the past year. Compared alongside Apple and other mapping technologies, Google is far ahead of the game.

Architectural detail on churches from Google maps, even the front steps are included. Photo: Justin O'Beirne.

Google is also using this data to create AOI's, Areas of Interest, on it's mapping service. Using it's building footprint data collected from satellites and the function of tho...

Pop quiz, hot shot: identify these world cities from above

2015-10-05T17:58:00-04:00

We’ve stripped out the street names and lost the labels – but can you still recognise the cities from their aerial views?

This exercise in aerial recognition comes in quiz form, where the viewer must guess the city pictured in a monochrome-treated satellite image of an urban grid. Identifying some cities is far easier than others – the quiz will tell you how your response stacks up against others'.