DocumentCode
3341290
Title
The Application of KL Transform to Remove Direct Wave in Ground Penetrating Radar Records
Author
Zhihua, Huo ; Minghui, Wang
Author_Institution
Sichuan Univ., Chengdu
fYear
2007
fDate
22-24 Aug. 2007
Firstpage
133
Lastpage
138
Abstract
In ground penetrating radar (GPR) records, because of different kinds of noises, especially direct wave of ground surface, the real target is not seen clearly in the radar echo section, which makes the detection of real targets difficult. Considering that the direct waves among traces are not change greatly and have strong correlation, so it can be restrained by certain methods. KL transform often be used to pick up coherent noises and random noises in multi channel seismic records[U 4]. In this paper, KL transform is used to GPR data processing, which fully uses the different second-order statistic characteristics of their correlations on the direct wave, target signals in multi channel GPR data. The target signals are successfully separated form noises according to the corresponding relation of the eigenvalue´s magnitude of the correlation matrix and correlation strength of each component in the multi channel GPR data.
Keywords
correlation methods; eigenvalues and eigenfunctions; ground penetrating radar; matrix algebra; radar cross-sections; radar detection; radar target recognition; radar tracking; random noise; transforms; KL transform; coherent noise; correlation matrix; direct wave removal; eigenvalue magnitude; ground penetrating radar record; multi channel GPR data; multi channel seismic record; radar echo section; random noise; real target detection; Application software; Computer graphics; Covariance matrix; Data processing; Filters; Ground penetrating radar; Karhunen-Loeve transforms; Signal processing; Statistics; Surface waves;
fLanguage
English
Publisher
ieee
Conference_Titel
Image and Graphics, 2007. ICIG 2007. Fourth International Conference on
Conference_Location
Sichuan
Print_ISBN
0-7695-2929-1
Type
conf
DOI
10.1109/ICIG.2007.173
Filename
4297069
Link To Document