Boundaries extraction and features recognition of metro station catchment area: the perspective of cognition and perception

Extracting boundaries and recognizing features of metro station catchment area (MSCA) are crucial for rational urban and transportation planning. Metro stations characterized by their high concentration of people and activities, serve as key locations for human-environment interaction and the development of cognitive relationships with individuals. Traditional research on MSCA has largely relied on distance, based estimations, often overlooking the cognitive influences. This study introduces a novel framework for the extraction of MSCA boundaries and the recognition of significant features. It examines the relationship between Points of Interest (POIs) and metro stations from two pivotal dimensions: cognition and perception. Cognitive relevance is evaluated through web co-occurrence analyses and combined effects, whereas perceptual relevance is assessed by simulating pedestrian perceptions using a cumulative Gaussian function. Utilizing the highly relevant POIs and road network data, the Alpha shape algorithm is deployed and combined with road data to delineate boundaries. Furthermore, Weighted Kernel Density Estimation is employed to recognize features and their distribution. The results show that our framework can well represent MSCA boundaries and features. Such an approach enhances the understanding of human-environment interactions in metro-centric areas and offers a data-driven approach for optimizing urban layouts, improving land use, and promoting transit-oriented development.