Detection and Localization of Object using Forward-Backward Time-Stepping for Through-wall Imaging Application
Through-wall imaging (TWI) is one of the largest microwave application nowadays. It is able to locate and reconstruct the image of an object hidden behind the walls. However, wall clutter in the TWI system is one of the main problem other than noise. It causes distortion and erroneous image reconstr...
Saved in:
| 主要作者: | |
|---|---|
| 格式: | Thesis |
| 語言: | English |
| 出版: |
2019
|
| 主題: | |
| 在線閱讀: | http://ir.unimas.my/id/eprint/31492/4/Mohamad%20Faizal%20full.pdf |
| 標簽: |
添加標簽
沒有標簽, 成為第一個標記此記錄!
|
| 總結: | Through-wall imaging (TWI) is one of the largest microwave application nowadays. It is able to locate and reconstruct the image of an object hidden behind the walls. However, wall clutter in the TWI system is one of the main problem other than noise. It causes distortion and erroneous image reconstruction of a hidden object. In this research, the forward-backward time stepping (FBTS) technique has been employed and integrated with Wall Direct Subtraction (WDS) to mitigate the wall clutter. Wall clutter are effectively reduced by the proposed method for through homogenous (99.59%) and heterogeneous (94.62%) walls, in medium, respectively. The work has included well reconstruct images of the hidden object with different shape, size and location. When Gaussian random noise added in the TWI simulation setup, the image reconstruction of an object would show severe distortion in its shape and composition. Thus, the Singular Value Decomposition (SVD) based filter specifically Singular Value Elimination and Savitzky-Golay filter (SVE-SG) method are used to reduce the noise. SVE-SG performed in time-domain to avoid data shifting and frequency domain conversion. The performance of SVE-SG method is shown with lower mean square error (MSE) values. It provides improvement from 22.65% to 83.00% when SVE-SG method applied through different wall thickness and electrical properties. At different noise levels, SVE-SG improves noise reduction in the range of 3.25% to 42.23%. |
|---|