The removal of additional edges in the edge detection of potential field data

Edge detection results of potential field data are used to delineate the horizontal locations of the causative sources, and there are many edge detection filters to finish this work. However, most of balanced edge detection filters produce additional edges which interpret the potential field data that contain positive and negative anomalies. First, we test the application effect of several common balanced edge detection filters, and then analyze the reason that produces additional edges. We present several new edge detection filters depending on the distribution features of different derivatives that will not produce additional edges. The new filters are demonstrated on synthetic gravity anomalies, which show the edges more precisely, and are insensitive to noise. We also apply them to real potential field data because they display the locations of the stratigraphic markers more precisely and clearly.