Towards the Automated Mapping of Linear Anomalies within Aeromagnetic Datasets

Aeromagnetic datasets can be used for regional- scale geological mapping in poorly exposed or studied geological terrains. Linear anomalies in aeromagnetic images correspond to faults and dykes, and their discontinuities and change of orientation have significant impact on various mineralisation phenomena. Thus for mineral exploration, robust detection and mapping of linear anomalies are an important step in interpreting aeromagnetic data. This paper presents an automatic tool that maps linear anomalies in aeromagnetic datasets. It detects dominant linear anomalies and then identifies their skeletal segments, revealing offsets within the linear anomaly. This is achieved through the following processes. Firstly, the regions of magnetic discontinuity corresponding with lithological boundaries and crustal breaks are identified through texture and bilateral symmetric feature detection. Secondly, dominant linear anomalies are identified using the Radon transform from the regions of discontinuity output from the first stage. Thirdly, we detect the skeletal segments that belong to each of the dominant linear features. This is achieved by finding all skeletal line segments of the regions of discontinuity and identifying the line segments that correspond with each of the dominant linear features using their orientations and locations. Finally, the skeleton segments for each of the linear anomalies are overlaid on the aeromagnetic image. An experiment was conducted using aeromagnetic data of the Yilgarn Craton in Western Australia.