TR2018-098
Multi-Layer Terahertz Imaging of Non-Overlapping Contents
-
- "Multi-Layer Terahertz Imaging of Non-Overlapping Contents", IEEE Sensor Array and Multi-Channel Signal Processing Workshop (IEEE SAM), DOI: 10.1109/SAM.2018.8448438, July 2018, pp. 652-656.BibTeX TR2018-098 PDF
- @inproceedings{Wang2018jul2,
- author = {Wang, Pu and Fu, Haoyu and Koike-Akino, Toshiaki and Orlik, Philip V.},
- title = {Multi-Layer Terahertz Imaging of Non-Overlapping Contents},
- booktitle = {IEEE Sensor Array and Multi-Channel Signal Processing Workshop (IEEE SAM)},
- year = 2018,
- pages = {652--656},
- month = jul,
- doi = {10.1109/SAM.2018.8448438},
- url = {https://www.merl.com/publications/TR2018-098}
- }
,
- "Multi-Layer Terahertz Imaging of Non-Overlapping Contents", IEEE Sensor Array and Multi-Channel Signal Processing Workshop (IEEE SAM), DOI: 10.1109/SAM.2018.8448438, July 2018, pp. 652-656.
-
MERL Contacts:
-
Research Area:
Abstract:
This paper considers terahertz (THz) imaging of multi-layer nonoverlapping contents with compressed measurements. One issue here is the shadow effect from front layers to deep layers due to the non-uniform penetrating illumination. In the case of nonoverlapping contents in layered structures, the shadow effect can be utilized to improve recovery performance and reduce the number of measurements. To this end, we propose several approaches based on the total variation (TV) minimization principle and take into account individual-layer sparsity, group sparsity over layers, and hierarchical group sparsity over layers to reduce the number of measurements. Numerical evaluation confirms the effectiveness of the proposed approaches.