Open AccessMulti‐feature enhancement for terahertz imaging
Author(s) -
Wang Tianhe,
Ding Jinshan,
Zhang Yuhong
Publication year - 2019
Publication title -
electronics letters
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.375
H-Index - 146
ISSN - 1350-911X
DOI - 10.1049/el.2019.1962
Subject(s) - terahertz radiation , feature (linguistics) , artificial intelligence , image quality , convolution (computer science) , computer vision , computer science , image (mathematics) , image resolution , noise (video) , set (abstract data type) , pixel , optics , pattern recognition (psychology) , physics , philosophy , linguistics , artificial neural network , programming language
A terahertz (THz) imaging enhancement approach is presented by using multi‐features of images. First, a set of THz images are reconstructed from a single data collection using THz time‐domain spectroscopy system by selecting different features. Second, the THz beam model is considered to improve image resolution, which is realised via a Wiener de‐convolution operation. Third, an approach to suppress the noise and enhance image feature is adopted to achieve a better image quality. Finally, three high‐quality images are selected from the new image set as red, green, and blue channels, respectively, to form a pseudo‐colour image. The experimental results suggest that the pseudo‐colour images show multiple features of the sample and achieve a better image quality.