Magnetic Anomaly Detection Based on Robust Principal Component Analysis and Orthonormal Basis Functions

Magnetic anomaly detection is a technology that identifies weak target magnetic signals in complex background noise to locate ferromagnetic objects. In practical applications, this background noise often consists of complex colored noise, necessitating effective suppression prior to the analysis of magnetic measurements. The measured magnetic field data usually consists of a low-rank background and sparse target signals. Since the target signal exhibits isolated outlier characteristics, it can be separated from the background into sparse components. This paper introduces the robust principal component analysis (RPCA) method to decompose and reconstruct the detection data in order to improve the SNR. First, the one-dimensional signal is converted into a block model matrix, and the matrix is decomposed into a low-rank component and a sparse component using the enhanced Lagrange multiplier method (ALM). The sparse component is then inversely transformed and reconstructed into a high SNR signal. The reconstructed signal is then detected using an orthogonal basis function (OBF) with strong anti-white noise capability. Experimental results using simulated data and measured data show that this method can improve the detection performance compared with the original OBFs method.