Prestack Seismic Signals Wiring Harness Radial Median Filter

Improving both the signal-noise ratio and resolutions of seismic data is the vital subject in seismic explore field. In recent ten years, the main purpose of the field and indoor technology is to outstand the effective wave and suppress the interference wave, and to transform some interference wave to useful signal from which we can obtain some information which can describe the underground geologic message effectively. So, how to suppress the interference wave in order to outstand the effective wave is important. In this article, several normal non-linear noise elimination filters are introduced and their advantages and disadvantages are enumerated. On the base of the signal-noise separation condition of the seismic reflection in near surface engineering geophysics, we study the multi-radial mid-value filtering method which applies for processing pre-stack trace gather.As mid-value filtering is non-linear, and its mathematics theory is complicated, this paper only compares with the linear filter in their statistical character. The multi-radial mid-value filtering makes the considered interference waves such as surface wave, refract wave, direct wave to be separated or extracted as useful signal. The separated wave-field record is helpful to study the structure of the near surface, providing parameter information for static with low price. At the same time, the valid separation of these linear interference wave increases the signal-noise ratio and resolutions of the record which only aims at obtain the reflect wave information. Besides, the studying subject in this article, not only adapts to filtering in any linear direction, but also can filter in limited zone(such as fan-shaped, triangle etc).The feasibility and effective of the multi-radial mid-value filtering have been verified through some model data tests and suppressing the surface wave in the practical pre-stack seismic data. In addition, good effect has achieved from the separation of up or down wave for the practical VSP data. To sum up, the multi-radial mid-value filtering can separate the linear noise any direction effectively, and has the characters such as the high signal-noise separation, highly keeping amplitude and high resolution.