A bilateral optimization iteration CT reconstruction method for incomplete projections. (May 2022)

• Record Type: Journal Article
• Authors: Xiao, Yu; Zhong, Wang; Xu, Feng; Li, Yongcun; Hu, Xiaofang
• Is Part Of: Optics and lasers in engineering. Volume 152(2022)
• Publication Date: 2022-05
• Subjects: Hotomechanics -- Synchrotron radiation computed tomography -- Incomplete projection reconstruction; Lasers in engineering -- Periodicals; Optical measurements -- Periodicals; Optics -- Periodicals; Lasers en ingénierie -- Périodiques; Mesures optiques -- Périodiques; Optique -- Périodiques; 621.36605
• DOI: 10.1016/j.optlaseng.2021.106867
• Languages: English
• ISSNs: 0143-8166

Abstract

Highlights: A new incomplete projection reconstruction algorithm was proposed based on the discussion of key points in iteration, rather than some specific reconstruction method. The iteration process was improved from the two aspects of initial value and restraint equations. The reconstructed quality of new method was better than conventional methods for both model image and actual experimental image. The particle in local region was reconstructed better while it was wrongly recognized as two small particles using conventional methods. Abstract: A new incomplete projections CT reconstruction method was proposed by considering the key points of iteration process. The limitations of two key points in iteration method were discussed. The new method was obtained by discussing the iteration process of the two methods, which was called bilateral optimization iteration reconstruction method. By providing more restrain equations of projection in iteration, the sinogram TVM process could optimize the ART process to obtain more accurate reconstruction results. And by providing more accurate initial value for iteration, the ART method could optimize the sinogram repairment by TVM process to obtain more accurate projection information as initial values. A series of numerical reconstruction experiments were conducted to validate the new method. The overall and detailed quality of model image reconstructed by new method was better than that of conventional methods. And the overall and detailed quality of actual experimental images reconstructed by the new method was also better than conventional methods.