Optimization of unmanned aerial vehicle application for measuring in complex urban green space with rich landscape value

Abstract

The main tree biometric variables of urban green space can derive indicators of multiple ecological benefit values. Compared with plantations and natural forests, urban green space is usually featured with higher tree richness, more complex landscape design, and a higher degree of spatial heterogeneity. Appropriate designs for flight variable settings in complex urban space can significantly improve the quality of data, and bring large differences in economic and time costs in practical applications. However, current studies have a limited understanding of flight variables setting for UAV applications in complex urban green spaces. We carried out rigorous and practical designs for flight variable settings with high-frequency UAV oblique image measurements. We tested and quantified the effect of the settings of main flight variables (i.e., flight altitude, image overlap, gimbal pitch and solar elevation) on the process data and final data quality in a representative complex urban green space. In the limited range of flight variable settings, the data accuracy is significantly different in the complex urban green space. We found that solar elevation is the most important factor affecting the data quality. Specifically, flight operations with a lower solar elevation provided a higher measurement accuracy and outstanding cost-effectiveness which is significantly different from the solar elevation setting for measuring in plantations. Our results (which are different from the previous study on homogenous green spaces) also suggested to avoid strong sunlight during flight operations and to ensure that the flight altitude is three to four times higher than the tallest tree at least to improve data quality. The settings of the image overlap and the gimbal pitch should take more time cost and economic benefits into consideration. This study demonstrated the feasibility of UAV applications in complex urban green spaces, and the impact and importance of each flight variable on the dataset quality. The systematic analysis, quantitative results, and exact visualizations can facilitate a wider application space (i.e., urban forestry, landscape ecology, urban planning and urban design) and higher application quality of UAV oblique in urban green space.