Seismic Data Based On Lifting Wavelet Transform Lossy Compression Studies

Seismic exploration is a significant way to get information inside the earth, and it is also the basic way to find minerals and other resources. Today, seismic exploration technique is developing in the direction of multidimensional, multicomponent, multiparameter and high resolution, so that the seismic exploration data is increasing day by day. It becomes a serious problem to transform and store the data. Therefore, how to compress the data becomes the key to solve the problem.According to the analysis the character of seismic data, the data has redundancy information. The paper is based on wavelet transform, and studied the theory that how different wavelet bases convert to lifting scheme, so they can be used to transform seismic data. There are three classical compression method based on wavelet transform:EZW, SPIHT and EBCOT algorithm. The article studied each theory and the advantages and disadvantages, so the article chose SPIHT as the coding algorithm for seismic data. As for the original seismic data and the constructed data, the article designed five different evaluation criterions to analyze the constructed data. The experimental result shows:(1). The effect of the constructed data is different on the basis of different lifting wavelet transform. The db4 wavelet base and bior6.8 wavelet base have the best effect among all the chose wavelet bases.(2). According to the experimental result in the paper, transform the seismic data at the level of 5, and the compression rate is less than 10, the result of compression is best. Transform the seismic data at the level of 5, and the compression rate is between 10 and 20, the constructed result is the best.(3). With the SPIHT algorithm, the compression rate can be controlled precisely. When the compression rate is less than 5, the effect of the constructed data approximately equal to lossless compression, and when the compression rate increases, the quality of constructed data is reduced.In this article, the loss compression algorithm brings very good results, and the algorithm can be used to analyze different seismic data. Wavelet bases can be constructed according to different seismic data, and the SPIHT algorithm can be improved or using EBCOT algorithm to analyze seismic data.