Target recognition using scattering features extracted with EMD

Scattering features extracted via complex Empirical Mode Decomposition (EMD) are used for radar target recognition. EMD, which is associated with the Hilbert-Huang transform, can be used to extract inherent oscillations known as Implicit Mode Functions (IMFs) from a radar backscatter. These IMFs can be associated with the locations of the target scattering centers, and consequently with the target geometry or down range profile as seen by the interrogating radar. Features extracted using EMD are then presented to a distance-based target recognition system. The proposed target classification scheme is tested using synthetic and real radar data recorded in a compact range.