Removal of Ringing Noise in GPR data by Signal Processing

概要

論文の詳細を見る
Ringing is a common type of coherent noise in GPR data. When this type of coherent noise is strong and if we do not remove this noise effectively, deeper structure may be completely masked by this noise. Ringing appears as nearly horizontal periodic events, which is the most important feature enabling us to remove the noise by signal processing. Simple but frequently used processing techniques, such as background removal and residual filtering, are based on this appearance of the noise in GPR image. The methods based on Radon and wavelet transforms, relatively newer techniques, are also based on the same characteristics. Filtering in spatial wavenumber domain is also possible because the perfectly horizontal events correspond to a component having infinite wavelength in spatial wavenumber domain. In this paper, the processing techniques for removing ringing noise are reviewed and compared using field data. Furthermore, it is shown that ringing can be successfully filtered out by decomposing a GPR image into eigen-images using singular value decomposition and filtering the resultant eigen-images. This comparative analysis shows that the refined techniques are definitely more effective in removing the ringing noise with less distortion of GPR signals than the simple processing approaches but each method has its own merits and limitations. Moreover, preserving the horizontal and linear events having geological meaning such as groundwater table can be possible only through a kind of selective or local filtering, since the horizontal periodic characteristics of ringing noise are the most important starting point for applying signal processing techniques to remove the noise.
社団法人電子情報通信学会の論文
2005-10-18

Removal of Ringing Noise in GPR data by Signal Processing

スポンサーリンク

概要

著者

関連論文

スポンサーリンク