Blind-wavelet Algorithm And The Application Of It On De-noising Of The Metal Mine Seismic Data

With geological prospecting develops in the direction of deep prospecting,traditional metal exploration cannot fully meet the requirements of deep prospecting.The vertical resolution of the weight and magnetic exploration method is low. Andmagnetic field intensity attenuates inversely proportional with the square of thedistance. There are serious deficiencies in the exploration of deep metal ore. Thedepth of direct current prospecting is too shallow, and the resolution ofelectromagnetic method in the deep places is generally lower. In addition, because themetal ore is buried deeply and will be affected by the intrusion of hydrothermal andthe fracture, geophysical anomalies caused by the electrical and magnetic parametersbetween the ore body and surrounding rock get quite weak, which makes thetraditional exploration method cannot adapt to it, and new method need to be explored.And the seismic exploration means has the advantages of detection distance, highresolution, and high accuracy.In general, the geological environment of metal mineralization region is rathercomplex, tectonic movement is also active. Interference received by its seismic data ismuch more than oil and gas seismic data, and even hyperbolic features cannot befound in the single shot record. For such seismic data with relatively stronginterference, low resolution, low signal to noise ratio, it is difficult for the traditionalmethod to eliminate the noise. Therefore, we need to design a new de-noisingalgorithm to suppress or eliminate the interference noise of metal ores seismic data.In the context of studying the metal mine seismic data de-noising, this papermainly focuses on the following aspects: this paper has done a research on thedistribution of seismic data noise of metal ores; it deeply studies some related theoriesabout the blind signal separation, and finds out the applicability, advantages anddisadvantages of the blind separation of the de-noising; it also makes a deep study onthe wavelet threshold de-noising theory, constructs threshold de-noising method underthe multi-dimension space respectively for the one-dimensional signal andtwo-dimensional signal, and finds out the advantages and disadvantages of thewavelet threshold de-noising; it forms the one-dimensional Blind-Wavelet algorithmand two-dimensional Blind-Wavelet algorithm to eliminate seismic data of metal mineaccording to the advantages and disadvantages of Blind-Wavelet separation andwavelet threshold de-nosing, which gives full play to the advantages of blindseparation de-noising and wavelet threshold de-noising, and overcomes thedeficiencies of both methods. In addition, it discusses the applicability and theselection of parameters of the Blind-Wavelet algorithm. And it also discusses theselection of algorithm parameters by using the actual seismic signals. In the final partof this paper, it applies the constructed Blind-Wavelet algorithm to deal with theactual metal seismic data, proving that Blind-Wavelet algorithm has the obvious de-noising advantage, which can effectively eliminate the interference of randomnoise, the linear noise, surface waves, sound waves and other noise. This methodmakes the de-noising seismic data hold a fairly obvious hyperbolic performance,clearer and more fluently seismic sections, and a higher ratio of signal to noise. Itmeets the requirements of high-resolution, high fidelity after the de-noising, and hasgreat theoretical and practical significance to the deep metal ores seismic exploration.