Magnetic Anomaly Detection Based on Gradient Tensor Minimum Entropy and Variational Modal Extraction
W. Zhan, Genzhai Peng, Chengdong Wang, Chengfei Peng, Feng Jie Zheng, Yong Chen · 2024
Magnetic Anomaly Detection (MAD) is a kind of covert passive method to detect ferromagnetic objects. Many MAD methods, such as Orthogonal Basis Function (OBF) which is very classical, are very dependent on some priori information. While in practical applications, the priori information is usually difficult to obtain. In this paper, a MAD method based on modal extraction and entropy detection of the full magnetic gradient signal is proposed, which does not require any priori information. Firstly, the noise of the magnetic gradient signals is suppressed using Variational Modal Extraction (VME). Then the Minimum Entropy Detector (MED) is designed to obtain the entropy of the five gradient components respectively. Finally, the total entropy of the signal is calculated out and regarded as the output of the anomaly detection. Compare to other methods, the method proposed in this paper does not rely on any priori information. Simulation and experimental results show that the method can improve the Signal-to-Noise Ratio (SNR) more obviously than other methods, and can detect magnetic anomalies efficiently at very low SNR.