NYC's Urban Textscape

Yufeng Zhao and Matt Daniels turn years of New York City Street View text into a visual tour of signs, slogans, storefronts, warnings, and neighborhood patterns. The result is part map essay, part word hunt, with the city doing most of the writing.

What is NYC's Urban Textscape?

The story starts with pizza—all 111,290 mapped matches—then uses familiar words to show how New York's visible vocabulary changes across the five boroughs. Each map can lead back to street images, so a neat geographic cluster can quickly become a row of awnings, parking notices, graffiti, delivery vans, or accidental poetry. Underneath the essay sits an OCR-built corpus drawn from millions of Street View panoramas. That scale makes tiny words feel civic-sized, while the image links keep the project attached to actual corners rather than floating away into pure statistics.

What you can do there

• Browse maps of words found on New York City streets
• Open matching Street View images
• Continue to the companion word-search tool

Why we picked it

Most city maps foreground roads, borders, and landmarks. This one treats the words pasted onto the city as terrain, revealing patterns that are easy to walk past and hard to see all at once. It rewards both careful map reading and shameless searches for odd phrases.

How to get the most from it

Begin with the featured maps and open a few image collections before deciding what a cluster means. Compare a citywide word with one tied to a neighborhood or business type. When you want to choose your own term, follow the companion All Text in NYC search tool and try a food, warning, brand, or local phrase.

Good to know

The underlying imagery spans 2007 through 2024, so the maps mix many moments rather than showing today's streets alone. The creators say coverage is strongest for English and other Latin-alphabet text. Treat the patterns as an exploratory time capsule, not a complete live inventory of New York.

Who made it, and when?

Yufeng Zhao and Matt Daniels is the credited creator or organization. The earliest supported launch date we found is July 2025.