Traffic Flow Registraton for Unmanned Aerial Vehicle Detection

Registration of detection data is to ensure the time-space consistency and effective fusion of air-ground traffic detection data. In this paper, we aim to propose a space-time registration method for the vehicle detection data of unmanned aerial vehicle (UAV). Considering temporal and spatial distribution characteristics of UAV detection, we define time effectiveness and two evaluation indicators for UAV data registration, namely absolute and relative effective times. These two indicators are used to characterize the quality of UAV detection data. Based on time effectiveness of UAV detection, space and time registration methods are developed. Data verification is conducted by Matlab programming with the holographic vehicle trajectory data of I-80 highway. Results of the case study show that, registration largely improves the data accuracy of UAV traffic flow detection data from 68.4% to 88.8%.