High‐Fidelity Continuous‐Wave Terahertz Computed Tomography Based on Intensity Denoising Method

The terahertz wave is typically defined as ranging from 0.1 to 10 THz, and it is penetrating to nonpolar and nonmetallic materials, such as plastics, clothing, and bones. These characteristics have thrust terahertz imaging techniques into the limelight. As an effective imaging method, terahertz computed tomography enables the visualization of a three‐dimensional object’s internal structure in the form of cross‐sectional images. The object’s cross section is scanned in the translation and rotation directions by a beam of radiation, and then the projection data are reconstructed by using a tomographic reconstruction algorithm. In this paper, continuous‐wave terahertz computed tomography is built and the cross‐sectional images of object are reconstructed by filtered back projection algorithm. When the number of the projection angles increases, the cross‐sectional images of the object contain more noise, making the cross‐sectional images appear fuzzy. On the other hand, when the translation interval increases during the data collection process, the cross‐sectional images exhibit serious degradation, and the real‐size value of the sample cannot be accurately measured. The present system obtains the cross‐sectional images, which puts forward the optimal value about the number of the projection angles and the translation interval.