Search papers, labs, and topics across Lattice.
This paper introduces SLIDER, a novel framework for aerial robot target search that integrates a local sliding map with sparse global history to enhance exploration in large-scale unknown environments. By employing a unique observation quality evaluation method and an incremental viewpoint clustering strategy, SLIDER significantly reduces computational demands while maintaining high efficiency in target detection and decision-making. Extensive simulations and real-world tests reveal that SLIDER surpasses existing methods in memory efficiency, decision latency, and overall search effectiveness.
SLIDER achieves superior target search efficiency while slashing memory usage and decision latency, redefining aerial exploration capabilities.
Efficient exploration and target search in large-scale unknown environments remain challenging for aerial robots due to the demands of broad spatial coverage, fine-grained perception, and real-time decision-making. This paper presents SLIDER, a lightweight and memory-efficient framework that avoids reliance on globally dense maps by combining a local sliding map with sparse global history information. A novel observation quality evaluation method is proposed, leveraging historical poses and sensor models to assess point cloud data in real-time, enabling efficient frontier detection. To support scalable and responsive planning, an incremental viewpoint clustering strategy dynamically adapts to local updates, significantly reducing the number of candidate targets and decreasing computational load. A sparse global topological map is incrementally maintained to assist global planning and cost evaluation. Extensive simulations and real-world experiments demonstrate that the proposed system outperforms state-of-the-art methods in memory usage, decision latency, and search efficiency.
