Joint frame rate adaptation and object recognition model selection for stabilized unmanned aerial vehicle surveillance

We propose an adaptive unmanned aerial vehicle (UAV)-assisted object recognition algorithm for urban surveillance scenarios. For UAV-assisted surveillance, UAVs are equipped with learning-based object recognition models and can collect surveillance image data. However, owing to the limitations of UAVs regarding power and computational resources, adaptive control must be performed accordingly. Therefore, we introduce a self-adaptive control strategy to maximize the time-averaged recognition performance subject to stability through a formulation based on Lyapunov optimization. Results from performance evaluations on real-world data demonstrate that the proposed algorithm achieves the desired performance improvements.

Automated summary

This paper proposes an adaptive unmanned aerial vehicle (UAV)-assisted object recognition algorithm for urban surveillance scenarios. By equipping UAVs with learning-based object recognition models and allowing them to collect surveillance image data, the proposed algorithm aims to maximize recognition performance while taking into account the limitations of UAVs in terms of power and computational resources. Through the introduction of a self-adaptive control strategy based on Lyapunov optimization, the algorithm achieves the desired performance improvements as demonstrated in real-world evaluations.