Data processing methods for terahertz transmitted spectral imaging

摘要

THz wave imaging is a new and promising technology for material distinguishing and non-destructive testing. The processing of the time domain waveform acquired at each pixel of the image is one of the most important aspects of the THz imaging. Due to the diversity of samples in the practical applications, signal processing approaches are quite different. A good method permits high quality of the images. In the previous papers, some methods have been adopted to generate the image of the object; however, these methods have not been compared, and the reason for choosing certain approach has not been given. Therefore, it is very important to discuss the difference of different imaging approaches for a sample and to choose the best one. In this presentation, several data processing methods for terahertz transmitted spectral imaging based on classical transmitted scanning imaging setup and the Fourier transform, including time-domain maximum-amplitude (minimum-amplitude) imaging, time-domain peak-to-peak amplitude imaging, time-domain time delay imaging, frequency-domain fixed frequency amplitude (power) imaging, as well as frequency-domain fixed frequency phase imaging, are discussed and compared. The results indicate that the imaging quality is different for different approaches; furthermore, the optimal imaging methods for the samples used in this paper are also different.