Efficient seismic compression using the lifting scheme

The Discrete Wavelet Transform (DWT) has a potential impact on the design of compression schemes and has been recently successfully applied to seismic compression. Some application areas such as the transmission of raw field data via satellite and mass storage do not put stringent limits on the computational complexity and on the extra storage space for the DWT coefficients. However, the situation becomes critical when the compressed data is devoted to the interpretation of the final product on workstations. In this contribution, a novel seismic compression scheme is proposed where the biorthogonal transform and its inverse are implemented via the lifting scheme. The data reduction is achieved by shrinking the wavelet coefficients using a scale-dependent non-linear soft thresholding. The adaptive threshold is derived using the Birge-Massart strategy. The proposed technique is applied to a migrated seismic section from the Midyan Basin (Red Sea, Saudi Arabia).