Effect assessment of spatial sampling length on the orthonormal basis functions detector

Abstract

Magnetic anomaly detection (MAD) is a passive way to detect hidden magnetic objects. As the major way of MAD, the orthonormal basis function (OBF) detector has been widely researched and applied. During the practical signal processing process, we found that the performance of the OBF detector is mainly affected by the space sampling length. However, there is almost no literature studying the relationship between the OBF detector and the space sampling length. In our work, the effect of the spatial sampling length on the performance of the OBF detector is assessed from the perspective of the statistical distribution of independent and correlated background noise. The theoretical and experimental results both turn out to be that the smaller the noise correlation, the more evident the gain effect of the spatial sampling length on the performance of the OBF detector under the same signal-to-noise ratio. Therefore, the spatial sampling length should be fitted with the noise environment to attain a better trade-off between detector performance and computing resources.