Magnetic anomaly detection (MAD) of ferromagnetic pipelines using principal component analysis (PCA)
Magnetic anomaly detection (MAD) of ferromagnetic pipelines using principal component analysis (PCA)
复制标题
DOI：
10.1088/0957-0233/27/4/045104
复制
发表时间：
2016-03
影响因子：
2.4
通讯作者：
A. Sheinker;M. Moldwin
中科院分区：
文献类型：
--
作者：
A. Sheinker;M. Moldwin
磁异常检测（MAD）方法用于检测视觉模糊的铁磁物体。该方法利用源自铁磁物体的磁场，该磁场构成周围地球磁场中的异常。传统上，MAD用于检测具有偶极结构的磁场的物体，其中远离物体的磁场可以被认为是点源。在目前的工作中，我们扩大MAD的情况下，一个非偶极子源，即一个铁磁管道。我们使用主成分分析（PCA）来计算主成分，然后使用这些主成分来构造有效的检测器。在我们的实验室进行的实验与真实世界的数据验证了上述分析。简单，低计算复杂度，和高检测率，使所提出的检测器有吸引力的实时，低功耗的应用。
The magnetic anomaly detection (MAD) method is used for detection of visually obscured ferromagnetic objects. The method exploits the magnetic field originating from the ferromagnetic object, which constitutes an anomaly in the ambient earth’s magnetic field. Traditionally, MAD is used to detect objects with a magnetic field of a dipole structure, where far from the object it can be considered as a point source. In the present work, we expand MAD to the case of a non-dipole source, i.e. a ferromagnetic pipeline. We use principal component analysis (PCA) to calculate the principal components, which are then employed to construct an effective detector. Experiments conducted in our lab with real-world data validate the above analysis. The simplicity, low computational complexity, and the high detection rate make the proposed detector attractive for real-time, low power applications.