关键词: 地震数据压缩, EZW, 小波变换, DCT, 分频

Abstract: With the significant improvement of high-density acquisition and daily acquisition efficiency, the amount of exploration data is growing exponentially. In order to save storage space and improve transmission efficiency, the research on high-fidelity compression and reconstruction of prestack seismic data is extremely urgent. EZW (embedded zerotree wavelet) algorithm has excellent compression performance. The traditional EZW algorithm for seismic data compression is based on data in the t-x domain. Because the wavenumber of prestack seismic data is wide and the frequency band is limited, the energy of the discrete cosine transform (DCT) data is mainly concentrated in the low-frequency in the DCT domain. This paper proposes a frequency division seismic data compression technology. Firstly, 2D-DCT is performed on the t-x data, and then frequency division is applied to the 2D-DCT data to obtain high-frequency and low-frequency data. By separately coding high-frequency and low-frequency data based on EZW, the low-frequency data has a high coding level to ensure the signal-to-noise ratio (SNR) of the reconstructed data, and the high-frequency data has a low coding level to improve the data compression ratio (CR). The synthetic data and field data experiments show that the compression ratio with the proposed method is higher than that with the conventional EZW at the same SNR.

Key words: seismic data compression, EZW, wavelet transform, DCT, frequency division