In this work, we present our architecture for automated optical tracking of a moving ground target from an airborne observer. After laying out the motivation of enabling stable and automated optical following of ground targets for airborne multi-target-tracking (MTT), we describe the implementation in detail. Building on the ROS2 framework, we use modularized object detection, target state estimation, and gimbal control modules which, depending on the current situation, use direct optical measurements or estimated measurements as controller input. Further, we show how parallel heterogeneous filters are applied to harness the benefits of integrating geographic information system (GIS) data in the form of road networks into the data fusion process while still keeping a second transition model for more general use. Using small-scale lab experiments and flight data collected earlier with a manned ultralight aircraft equipped with a state-of-the-art surveillance sensor, we verify and test the state estimation process and the gimbal controller. Along with these preliminary experiments, a planned flight campaign will demonstrate the integrated system in flight.
Airborne Optical Sensing: Improving Ground Target Tracking with GIS Road Network Integration
2024-09-29
828494 byte
Conference paper
Electronic Resource
English
Ground target tracking with road map support
Tema Archive | 2007
|Road-Map Assisted Ground Moving Target Tracking
Online Contents | 2006
|Road-map assisted ground moving target tracking
IEEE | 2006
|