多尺度分析方法可以在不同的尺度上将信号分解成为轮廓和细节，这样有助于对信号进行更为精细的处理。简要介绍了多尺度方法应用于重磁数据处理的基本思想，从理论上讨论了基于多尺度方法提高重磁场深／浅源异常对比度的方法，并对其作了严格的证明。理论分析及模型实例表明，多尺度方法虽然不能将性质相同的异常完全分离，但可以有效地提高深源异常对浅源异常的对比度。在有浅源异常干扰的地区，可以有效地压制浅部干扰，突出深部场源信息。 The multi-scale analysis method can decompose the signal into the outline and the detail on different scales, thus can help to carry on the finer treatment to the signal. The basic idea of ​​multi-scale method applied to gravity and magnetic data processing is briefly introduced. The method to improve the contrast of deep / shallow anomalies based on multi-scale method is theoretically discussed and strictly proved. Theoretical analysis and model examples show that although the multi-scale method can not completely separate the anomalies of the same nature, it can effectively improve the contrast of the deep anomalies to the shallow anomalies. In areas with light-source anomalies, it is possible to effectively suppress shallow interferences and highlight deep source information.