Edge identification is an important task in the interpretation of potential field data and can be used in the field of resoureedeteetion and so on, which is one of the hot topics in current research. For deep anomalous targets, the traditional edge identificationmethods are prone to blur, divergence, deformation and other problems. Most of the improved methods use high-order derivatives,although they can well identify the deep anomaly boundary, the caleulation is complieated and easy to be disturbed by noises. In order toimprove the above problems, proposes a stable and fast edge identification method, which only uses sign funetion and gravity magneticvertical derivative data, can identify the targels at diferent depths in a balanced manner, and avoids the caleulation of higher-orderderivatives. Through experiments and comparisons of theoretical models,the convenience,effectiveness, anti-oblique magnetizationability and anti-noise ability of the method were verified. At the same time, through the application of actual data, it was verified that themethod ean accurately and clearly distinguish abnormal objeets in a complex gravity and magnetie environment, which ean save a lot oftime and energy of relevant researchers, and has a good application prospect.