Generating citywide street cross-sections using aerial LiDAR and detailed street plan

Abstract

Precise information on spatial configurations of urban spaces, such as streets, is essential for investigating complex interrelationships between the quality of life, livability, and mobility. Street cross-sections can be central for such research as they depict the allocation of urban space by detailing the streetscape's layout and dimensions, including driveways, sidewalks, bikepaths, median strips, trees, adjoining buildings and open spaces, and shadows. However, creating an accurate description of the real world as a three-dimensional representation and translation into cross-sections is challenging due to the requirement of multiple data sources, such as road layouts with widths, height information, and, optionally, high-resolution aerial imagery. Without such datasets, the cross-section drawings are limited to manual measurements, which are difficult to extend to a city scale. This study aims to develop a method to automatically generate street cross-sections for the entire city of Berlin based on an aerial Lidar dataset and a city street plan. The study also includes shadows as a part of the sections and utilizes the Lidar dataset to generate solar radiation maps. Approximately 0.5 million cross-sections are generated with detailed information regarding the width and placement of the street elements. The details of the cross-sections are further processed to find generic street compositions, understand the influence of shadows on walkways and bikepaths, and calculate the enclosure.