EXPERIMENTALRESEARCHESONTHETHRESHOLDOFAIRBORNEGRAVITY DATADENOISINGBASEDONDBWAVELETTRANSFORM
-
摘要: 利用Mallat快速算法实现离散小波变换,选用DaubechiesN小波系为小波基函数,通过理论模型数据验证了方法设计的可靠性。在对实测航空重力数据进行去噪过程中,根据重力异常信号频带特点,采取多层分解和阈值策略进行自由空气重力异常的提取。试验进行了6~8层的分解,采用了强制阈值去噪和施加软阈值去噪,并进行了对比,结果表明基于DB小波阈值去噪所获得的自由空气重力异常与GT-1A系统滤波结果基本吻合。Abstract: ThispaperproposesamethodtocomputetheDWT(discretewavelettransform)coefficientsusingMallat'sfastalgorithmwith DaubechiesN(N=1,…,10)waveletseriesasthewaveletprimaryfunction.Theauthorsverifiedthereliabilityofthemethodwitha theoreticalmodel.Byusinganapproachtomulti-layerdecompositionandastrategyofthresholdvalueintheprocessoftheairborne gravitydatadenoising,theauthorsextractedthefreeairgravityanomaliesbythefrequencyandbandwidthcharacterofthegravitya-nomalies.Acomparisonismadebyusingsix,sevenandeightlayersdecompositionandbyemployingthesoft-thresholdandthehard-thresholddenoising.Theresultsdemonstratethatthefreeairgravityanomaliesextractedbythismethodarequiteconsistentwiththefil-teredresultsoftheGT-1Asystem.
-
Key words:
- DBwavelet /
- anomalydecomposition /
- thresholdselection /
- airbornegravity
-
-
计量
- 文章访问数: 436
- PDF下载数: 96
- 施引文献: 0