Google Street View has gathered millions of street view images for the past decade from across the U.S. Additionally, it keeps those maps regularly updated by re-photographing the places. This rich database has been used by researchers to follow cities through time.
Nikhil Naik and company have utilized an artificial intelligence algorithm to examine the images of some 1.6 million street blocks from five different cities. The algorithm calculates street change using pictures of the same streets captured seven years apart. A positive value of street change is generated for upgrades and new constructions, while a negative value for overall decline.
The study found that two demographic characteristic- high density and high education, play significant roles in urban improvement. The findings also supported three classical theories for urban change.
The first is the 'human capital agglomeration theory,' according to which urban improvement is witnessed in areas where there is a significant density of highly educated individuals.
The second is the ‘tipping’ theory, which argues that developed neighborhoods tend to develop further. And the third is the ‘invasion’ theory, according to which localities around successful neighborhoods tend to improve greatly over time.