Fast Lossless Compression of Seismic Floating-Point Data

Modern seismic exploration produces vast amount of data that may be transferred between computers as well as to and from storage devices. Compression algorithms are then desirable to make seismic data processing more efficient in terms of storage and transmission bandwidth. Most current algorithms being used for floating-point seismic data compression are based on wavelet or LCT (local cosine transform) and may result in information loss. In this paper, a new lossless floating seismic data compression method based on differential predictor and adaptive context-based arithmetic coding (DPACAC) is presented, which can achieve higher compression ratios compared with other lossless algorithms.