Cargando…

Mask Responses for Single-Pixel Terahertz Imaging

Terahertz (THz) radiation meaning electromagnetic radiation in the range from 0.1 (3) to 10 (30) has the unique advantage of easily penetrating many obstructions while being non-hazardous to organic tissue since it is non-ionizing. A shortcoming of this domain is the limited availability of high-sen...

Descripción completa

Detalles Bibliográficos
Autores principales: Augustin, Sven, Frohmann, Sven, Jung, Peter, Hübers, Heinz-Wilhelm
Formato: Online Artículo Texto
Lenguaje:English
Publicado: Nature Publishing Group UK 2018
Materias:
Acceso en línea:https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5861092/
https://www.ncbi.nlm.nih.gov/pubmed/29559708
http://dx.doi.org/10.1038/s41598-018-23313-6
Descripción
Sumario:Terahertz (THz) radiation meaning electromagnetic radiation in the range from 0.1 (3) to 10 (30) has the unique advantage of easily penetrating many obstructions while being non-hazardous to organic tissue since it is non-ionizing. A shortcoming of this domain is the limited availability of high-sensitivity detector arrays respective THz cameras with >1k pixels. To overcome the imaging limitations of the THz domain, compressive imaging in combination with an optically controllable THz spatial light modulator is a promising approach especially when used in a single-pixel imaging modality. The imaging fidelity, performance and speed of this approach depend crucially on the imaging patterns also called masks and their properties used in the imaging process. Therefore, in this paper, it is investigated how the image quality after reconstruction is specifically influenced by the different mask types and their properties in a compressive imaging modality. The evaluation uses an liquid-crystal display based projector as spatial light modulator to derive specific guidelines for the use of binary and true greyscale masks in THz single-pixel imaging setups respective THz single-pixel cameras when used in far-field applications e.g. stand-off security imaging.