Multi-scale Classification for Electrosensing

This paper introduces a premier and innovative (real-time) multi-scale method for target classification in electrosensing. The intent is that of mimicking the behavior weakly electric fish, which able to retrieve much more information about by approaching it. based on family transform-invariant shape descriptors computed from generalized polarization tensors (GPTs) reconstructed at multiple scales. evidence provided different each scale fused using Dempster--Shafer theory. Numerical simulations show recognition algorithm we propose performs undoubtedly well yields robust classification.