Change Detection of Urban Vegetation from Terrestrial Laser Scanning and Drone Photogrammetry

Urban areas experience continuous transformations, impacting the urban vegetation, particularly urban trees. The expansion of urban landscapes directly impacts green spaces and vegetation within cities. Urban vegetation plays a crucial role in improving the urban environment, benefiting residents' well-being, air quality, and temperature regulation. Monitoring changes in urban vegetation is therefore essential, considering the environmental and well-being aspects.

This study focuses on change detection using terrestrial laser scanning (TLS) and drone photogrammetry, utilizing three-dimensional (3D) point cloud data. Change detection compares multi-temporal datasets to analyze variations in a geographic region. TLS and drone photogrammetry techniques have gained popularity for monitoring urban vegetation, as they enable the acquisition of detailed 3D information. Point cloud data captures 3D information, enabling detailed change detection and 3D visualization of urban vegetation. This enhances the level of detail and information provided by the methodologies.

The objective is to estimate the growth of urban vegetation in a specific area within Helsinki's Malminkartano region during the spring and fall seasons of 2022 using multi-temporal TLS, UAV photogrammetry, and their integration. The research examines the suitability of different point cloud datasets acquired with different sensors and parameters for change detection analysis, identifying potential differences, challenges, and proposed solutions. Three distinct methods, namely C2C, C2M, and M3C2 are employed for point cloud comparison.

The results highlighted that manual processing is required to make the point cloud datasets comparable, with significant issues related to differences in point density and resolution. The sparser UAV photogrammetry datasets pose limitations on detailed analysis for change detection. The visual results reveal that TLS datasets detect changes in urban vegetation up to 2m, while UAV photogrammetry and integrated datasets up to 2.8m. However, applying a threshold at a 95% confidence level, 80-90% of significant changes in TLS datasets are observed up to 0.5m, up to 1m in UAV datasets, and up to 0.5m in integrated datasets. These changes represent the growth of urban vegetation during the leaf-off and leaf-on seasons examined. Overall, the utilized datasets provide valuable insights into changes in urban vegetation within the study area.