City Walk: Segmentation, evaluation, and mapping for sidewalk surface data

Urban surfaces significantly influence the social and environmental aspects of life in urban areas; however, cities lack records of their surfaces due to the expensive and time-consuming nature of the data collection process. The current research on sustainable urban surfaces is rather focused on their potential, and it is hindered by the scarcity of data, limiting our understanding of their current conditions, spatial distribution, and temporal evolution. Moreover, data management becomes increasingly challenging, particularly for cities with constant environmental, economic, and social transformations; hence the research will demonstrate the viability of an automated methodology.

The research will follow a multiphase pipeline that integrates recent advances in deep learning, computer vision, and geospatial processes for classifying urban surface materials and walkability using widely available orthographic and street-level images. The research will provide urban analysts and city agents with accurate and extendable methods to collect city surface data, which plays a vital role in addressing sustainability issues, including sustainable development and climate resilience.

48731/48732: MSSD Thesis
Architecture Department
Carnegie Mellon University – Pittsburgh
Fall 2022 – Spring 2023