Work

Semantic-Aware Observation Optimization for Multi-UAV Traffic Monitoring with IDM-Based Behavior Modeling

Year: 2026
Type: conference-paper
Abstract: This paper proposes a semantic-aware observation point selection framework for multi-UAV traffic monitoring systems. Unlike conventional geometry- or coverage-based approaches, the proposed method jointly exploits semantic information extracted from aerial imagery and traffic interaction characteristics derived from a Semantic-Aware Traffic Complexity Model (SA-TCM). A unified observation utility score is developed to capture spatial importance, semantic observation utility, traffic complexity, and energy cost. Based on this formulation, a constrained optimization problem is established to maximize the overall observation utility under UAV energy, capacity, and spatial separation constraints. A low-complexity greedy algorithm is further developed for efficient observation point selection and UAV deployment. Simulation results demonstrate that the proposed framework significantly improves observation utility and deployment efficiency compared with conventional distance-based and random strategies while maintaining low computational complexity.
Language: English
Open Access status: hybrid
OpenAlex ID: W7204087635
FWCI:
Cites: 20
Cited by:
Related to: 10