A Lossless Compression Method Of Seismic Data

With the rapid development of digital seismic observation technology, digital seismic data compression technique is widely applied for seismic data restore and transmission. The compression technique can not only enhance the storage capacity of recoding media, but also extend the frequency band of transmission system relatively. Without adding hardware, it can improve efficiency of present equipment and reduce costing of communication.This paper gives lossless compression algorithms for digital seismic data. The compression System composed of linear prediction and variable-size coder between different segments. In this method, coding and decoding use the symmetrical plans and the predictor with same construction. Coding method based on characteristic of data series and adjusts coding size between segments.Sampled data from a same system have same length, that is, they have same bits. There is less amplitude in noise than signal that we interested in a usual system, and output of this system give us some information we interested and, for coding, these data amplitude have relation and we can prediction them from other data before or after it. Because sampling rate is more than 2 times of the signal frequency, usually there is slow change of bit-length between sampled datum, and there is same bits in a long range.We get a linear prediction model based on the theory of spectrum analysis. According this theory a p-order linear predictor is like a p-order AR model and we can get prediction coefficients from the model that based on the imputed data, and then get difference between real data and predicted data.We just deal with err(difference) in coding, and based on the trait of serial errs we get different subsections and can accommodate the most errs if it can in one subsection, that means it can reduce the redundancy in errs. We disport a large number into two pieces and coding them, which can work well when there are many peak values.This algorithms was tested and can reduce abundance in data effectively and it has good performance in compressing.