EMND-SS: Equivalent Magnetic Node-Driven Sparse Surveying for Reliable Magnetic Mapping in Unknown Environments

The construction of traditional magnetic map is time-consuming and labor-intensive, and lacks guidance for lightweight magnetic surveying theory, resulting in unreliable magnetic maps reconstruction based on sparse surveying in unknown areas. To address this issue, this paper introduces an equivalent magnetic node-driven sparse surveying (EMND-SS) technique for reliable magnetic mapping. An integrated model is established to characterize the fluctuation, topology, quality, and quantity of equivalent magnetic nodes. Based on local magnetic field variations, the effective horizontal and vertical boundaries of the nodes are determined, enabling adaptive adjustment of the aircraft’s speed and cross-track surveying spacing. Under the requirement of low-complexity trajectory design, a quasi-spiral trajectory is adopted to perform non-uniform sparse surveying from outer to inner regions. This approach ensures the acquisition of high-quality, well-topologized magnetic nodes under constrained and sparse surveying conditions. Extensive results from multiple datasets and multimodal experiments show that the proposed EMND-SS technique is insensitive to both the sparsity of magnetic nodes and the choice of reconstruction methods. It achieves reliable reconstruction results, with both the magnetic feature accuracy and the position stamp accuracy reaching over 93 % of their respective ideal upper bounds, while the precision of reconstructed magnetic features exceeds 0.15 nT.