Terahertz digital holography image processing based on MAP algorithm

Abstract

Terahertz digital holography combines the terahertz technology and digital holography technology at present, fully exploits the advantages in both of them. Unfortunately, the quality of terahertz digital holography reconstruction images is gravely harmed by speckle noise which hinders the popularization of this technology. In this paper, the maximum a posterior estimation (MAP) filter is harnessed for the restoration of the digital reconstruction images. The filtering results are compared with images filtered by Wiener Filter and conventional frequency-domain filters from both subjective and objective perspectives. As for objective assessment, we adopted speckle index (SPKI) and edge preserving index (EPI) to quantitate the quality of images. In this paper, Canny edge detector is also used to outline the target in original and reconstruction images, which then act as an important role in the evaluation of filter performance. All the analysis indicate that maximum a posterior estimation filtering algorithm performs superiorly compared with the other two competitors in this paper and has enhanced the terahertz digital holography reconstruction images to a certain degree, allowing for a more accurate boundary identification.