A method of ground penetrating radar image denoising based on shearlet transform

Abstract

The invention discloses a ground-penetrating radar image denoising method based on shearlet transform. Firstly, frequency-domain Fourier transform is performed on collected ground-penetrating radar data, and then frequency-domain windowing processing is performed. The ground radar data is subjected to non-subsampling shearlet transformation, and then the shearlet coefficients of different scales and directions are subjected to threshold shrinkage processing using a bivariate model based on maximum a posteriori probability estimation, the coefficients are updated, and the inverse shearlet transformation is performed on the new shearlet coefficients. The ground penetrating radar frequency domain image after denoising is obtained. The algorithm is simple to implement, has a good denoising effect, and has good engineering practical value.

Description

technical field [0001] The invention belongs to the technical field of ground penetrating radar image denoising, and in particular relates to a ground penetrating radar image denoising method based on shearlet transformation. Background technique [0002] In geological exploration, ground penetrating radar is more and more widely used. Since there are targets with different physical properties underground, this discontinuity becomes the basis for radar technology to effectively detect underground targets. Compared with other geological surveys, Ground-penetrating radar technology has great advantages in the exploration of shallow underground targets. It not only has high resolution, but also does not cause damage to the detection site. It has high detection efficiency and fast detection speed. These advantages make ground penetration Radar technology plays an important role in the exploration of shallow underground targets. Its application fields involve underground polluti...

AI Technical Summary

Problems solved by technology

However, wavelet analysis itself has limitations

Wavelet is the optimal basis for objective functions with point-like singularities. The wavelet coefficients are sparse when analyzing such objects. However, the excellent characteristics of wavelet analysis in processing one-dimensional signals cannot be simply extended to two-dimensional or higher dimension

This is because the two-dimensional separable wavelet formed by the one-dimensional wavelet has only three limited directions: horizontal, vertical and diagonal. In the case of high dimensions, it cannot make full use of the unique geometric characteristics of the data itself, and cannot optimally Represents higher-dimensional functions with line or surface singularities

In fact, the direction information of two-dimensional images is often reflected in the singularity of straight lines or curves, rather than just point singularity. When analyzing such signals, wavelet cannot mine the direction information well.

Images