Image Reconstruction Based on Combination of Inverse Scattering Technique and Total Variation Regularization Method
In this research, Total Variation (TV) regularization method was incorporated with the Forward-Backward Time-Stepping (FBTS) algorithm to deal with the ill-posedness of the inverse scattering problem in the time domain. The effectiveness between FBTS without and with regularization method is compare...
Saved in:
| Main Author: | |
|---|---|
| Format: | Thesis |
| Language: | English |
| Published: |
2020
|
| Subjects: | |
| Online Access: | http://ir.unimas.my/id/eprint/30058/4/Image%20Reconstruction%20Based%20on%20Combination%20of%20Inverse%20Scattering.pdf |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | In this research, Total Variation (TV) regularization method was incorporated with the Forward-Backward Time-Stepping (FBTS) algorithm to deal with the ill-posedness of the inverse scattering problem in the time domain. The effectiveness between FBTS without and with regularization method is compared and analyzed by numerical simulations and calculation of Mean Square Error (MSE). Finite-Difference Time-Domain (FDTD) scheme is used to calculate the inverse scattering signals in forward time-stepping and adjoint field in backward time-stepping to reconstruct the microwave properties. The Forward-Backward Time-Stepping - Total Variation (FBTS-TV) regularization algorithm is in a two-dimensional case and implemented in C++ language executed in single computing. The FBTS-TV regularization method shows a good performance of reconstructing relative permittivity and conductivity profiles of the unknown embedded object for its size, shape, and location. The image reconstruction in enhanced by smoothing irregular contours while preserved the edges, and hence produced a better estimation of the image’s boundaries. A distinct improvement is shown in the reconstruction of the object’s relative permittivity. In the case of reconstruction of a simple object for relative permittivity, FBTS-TV improved the FBTS algorithm by 15%. |
|---|