By 2050, more than two-thirds of the global population is expected to live in cities. As urbanization increases, the quality of the urban environment becomes crucial for human well-being and sustainable development. However, measuring and monitoring urban quality is challenging due to the vast amount of data required. To address this issue, researchers from the University of Notre Dame and Stanford University developed a scalable method using machine learning to measure urban decay at a detailed level over time.
The researchers used the YOLOv5 model, an artificial intelligence system, to detect objects associated with urban decay. These objects included potholes, graffiti, garbage, tents, broken windows, discolored facades, weeds, and utility markings. They applied their method in three cities: San Francisco, Mexico City, and South Bend, Indiana. The model performed well in detecting these objects in heavily populated areas like San Francisco but struggled in less dense suburban areas like South Bend.
The findings showed that the trained model could effectively detect urban decay objects in different cities and neighborhoods. It provided valuable information to assess temporal and geographic variations in issues such as homelessness. The approach has the potential to be scaled and used to track urban quality and change across various cities and countries where street view imagery is available.
The researchers believe that this method can offer more efficient data collection compared to traditional economic sources. It can inform urban policy and planning and address social issues like homelessness. The model could be valuable for governments, non-governmental organizations, and the public in understanding and improving urban spaces for a better future.
>>Join our Facebook Group be part of community. <<