Search

The human spatial perception of urban streets has a high complexity and traditional research methods often focus on access surveys of human perception. Urban streets serve as both a direct conduit for pedestrians’ impressions of a city and a reflection of the spatial quality of that city. Street-view images can provide a large amount of primary data for the image semantic segmentation technique. Deep learning techniques were used in this study to collect the boring, beautiful, depressing, lively, safe, and wealthy perception scores of street spaces based on these images. Then, the spatial pattern of urban street-space quality perception was analyzed by global Moran’s I and GIS hotspot analyses. The findings demonstrate that various urban facilities affect street quality perception in different ways and that the strength of an influencing factor’s influence varies depending on its geographical location. The results of the influencing factors reveal the difference in the degree of influence of positive and negative influencing factors on various perceptions of the visual dimension of pedestrians. The primary contribution of this study is that it reduces the potential bias of a single data source by using multi-dimensional impact analysis to explain the relationship between urban street perception and urban facilities and visual elements. The study’s findings offer direction for high-quality urban development as well as advice for urban planning and enhanced design.

The visual quality and spatial distribution of architectural styles represent a city’s image, influence inhabitants’ living conditions, and may have positive or negative social consequences which are critical to urban sensing and designing. Conventional methods of identifying architectural styles rely on human labor and are frequently time-consuming, inefficient, and subjective in judgment. These issues significantly affect the large-scale management of urban architectural styles. Fortunately, deep learning models have robust feature expression abilities for images and have achieved highly competitive results in object detection in recent years. They provide a new approach to supporting traditional architectural style recognition. Therefore, this paper summarizes 22 architectural styles in a study area which could be used to define and describe urban architectural styles in most Chinese urban areas. Then, this paper introduced a Faster-RCNN general framework of architectural style classification with a VGG-16 backbone network, which is the first machine learning approach to identifying architectural styles in Chinese cities. Finally, this paper introduces an approach to constructing an urban architectural style dataset by mapping the identified architectural style through continuous street view imagery and vector map data from a top-down building contour map. The experimental results show that the architectural style dataset created had a precision of 57.8%, a recall rate of 80.91%, and an F1 score of 0.634. This dataset can, to a certain extent, reflect the geographical distribution characteristics of a wide variety of urban architectural styles. The proposed approach could support urban design to improve a city’s image.

With the emergence of various types of indoor positioning technologies (e.g., radio-frequency identification, Wi-Fi, and iBeacon), how to rapidly retrieve indoor cells and moving objects has become a key factor that limits those indoor applications. Euclidean distance-based indexing techniques for outdoor moving objects cannot be used in indoor spaces due to the existence of indoor obstructions (e.g., walls). In addition, currently, the indexing of indoor moving objects is mainly based on space-related query and less frequently on semantic query. To address these two issues, the present study proposes a multi-floor adjacency cell and semantic-based index (MACSI). By integrating the indoor cellular space with the semantic space, the MACSI subdivides open cells (e.g., hallways and lobbies) using space syntax and optimizes the adjacency distances between three-dimensionally connected cells (e.g., elevators and stairs) based on the caloric cost that extends single floor indoor space to three dimensional indoor space. Moreover, based on the needs of semantic query, this study also proposes a multi-granularity indoor semantic hierarchy tree and establishes semantic trajectories. Extensive simulation and real-data experiments show that—compared with the indoor trajectories delta tree (ITD-tree) and the semantic-based index (SI)—the MACSI produces more reliable query results with significantly higher semantic query and update efficiencies; has superior semantic expansion capability; and supports multi-granularity complex semantic queries.