Question 1

How did Google Street View transition from a prototype to a global tool?

Accepted Answer

The project began with a GIS expert building a prototype in just two weeks that featured 360-degree video captured from a camera hanging under a helicopter. Before it became a polished product, Google co-founder Larry Page personally drove around the San Francisco Bay Area with a camcorder to test if building facade footage would be useful. By 2007, the first official fleet consisted of SUVs outfitted with cameras, lasers, and GPS units, with trunks full of computers to prove that users wanted to see a street before arriving at their destination.

Question 2

What kind of specialized equipment has been used to map areas inaccessible to cars?

Accepted Answer

Google has adapted its technology to various environments using a diverse range of "vehicles." For snowy terrain like the 2010 Winter Olympics or ski resorts, they used modified snowmobiles. To navigate pedestrian zones and narrow urban areas like Stonehenge, they developed the "Trike," a three-wheeled bicycle. For hiking trails and desert dunes, they utilized the "Trekker" backpack and even strapped cameras to camels. Additionally, they used underwater apparatuses to map the Great Barrier Reef and a pushcart called the "Trolley" to capture the interiors of museums and the International Space Station.

Question 3

How does Google Street View protect the privacy of individuals captured in images?

Accepted Answer

Before any imagery is made public, it undergoes a compute-intensive processing step to detect and blur faces and license plates. This system has been in development since 2008 and currently uses a sophisticated combination of a "sliding-window" detector and a fast post-processor to ensure high accuracy. The automated system successfully blurs approximately 89% of faces and 94% of license plates. In cases where the technology misses a detail or over-blurs an object, users can flag the image for manual review by Google’s team.

Question 4

How does the "Smart Navigation" feature create a 3D feel in a 2D photo?

Accepted Answer

Street View creates an immersive experience by building a low-resolution depth map for every panorama. Newer imagery uses laser scanners to measure the distance to surfaces like buildings and roads, creating a 3D mesh. For older imagery captured without lasers, Google uses a computer vision technique called "optical flow," which analyzes how objects move across the screen as the car drives by to calculate depth. When a user moves their cursor, the app stretches the 2D photos over this 3D model, making it feel like the user is walking through a physical space rather than just looking at a flat image.

Question 5

What is "Immersive View" and how does it differ from standard Street View?

Accepted Answer

Immersive View is a living 3D model of a city rather than a static photo. It uses AI and "neural radiance fields" to fuse billions of Street View and aerial images into a volumetric representation. This technology allows the map to simulate real-time conditions, such as changing the lighting based on the actual weather and time of day or showing the current "busyness" of a restaurant. It also enables "Immersive View for routes," which allows users to fly through their entire planned journey in a 3D environment before they start driving.

Google Maps History and the Tech Behind Street View

How This Personalized Podcast Was Made

More like this