A Row Displacement Correction Algorithm for High-speed and Accurate Terahertz Raster Scanning Imaging

Abstract

Terahertz time-domain spectroscopic imaging has become a common and useful technique for various interdisciplinary studies. Raster scan is still the most general and reliable imaging modality due to the lack of matrix detectors. In terahertz raster scanning imaging, row displacements occur due to the use of a two-way continuous scanning mode, the lack of delay line triggering, or other technical factors. The displacements result in image distortion which significantly affects the subsequent data analysis. In this article, a row displacement correction algorithm based on a genetic algorithm is proposed. The algorithm corrects images based on the principle that row displacements result in an increase of high-frequency components of the Fourier spectrum transformed along the image column direction. The algorithm performance is comprehensively verified by both numerical simulation and experimental data, showing a promising effectiveness and robustness on different sample types, shapes, orientations, and spatial distributions. The algorithm corrects the displacements numerically regardless of the causing factors, offering a wide adaptability for various THz systems to improve the scanning efficiency and image quality.