Image Reconstruction Based on Total Variation Minimization for Radioactive Wastes Tomographic Gamma Scanning From Sparse Projections
Rui Shi,Honglong Zheng,Xianguo Tuo,Changming Wang,Jianbo Yang,Yi Cheng,Mingzhe Liu,Songbai Zhang
DOI: https://doi.org/10.1109/access.2021.3088746
IF: 3.9
2021-01-01
IEEE Access
Abstract:Tomographic Gamma Scanning (TGS) is one of the most important non-destructive analyzed techniques for radioactive waste drums. By reconstructing the radioactivity distribution image, it can accurately realize the qualitative, quantitative, and positioning analysis of the radionuclides in the drum. However, the time consuming of the scanning is long and the reconstructed image is rough, which limits its good application in the practical assay of the waste drum. In this work, the total variational minimization (TVM) method was applied to improve the iterative process of the conventional algorithms of maximum likelihood expectation maximization (MLEM) and algebraic reconstruction technique (ART), then the MLEM-TVM and ART-TVM reconstruction methods were developed. The transmitted experiments were carried out where four kinds of materials were arranged in a segment whose densities ranging from 1.04 g/cm<sup>3</sup> to 2.02 g/cm<sup>3</sup> and a <sup>152</sup>Eu isotope was set up as a transmission source. Compared with the traditional algorithms MLEM and ART, the MLEM-TVM and the ART-TVM algorithms have a better performance on the accuracy and the signal-to-noise ratio, and the MLEM-TVM algorithm achieves the best results, which means the quality of the reconstructed image is improved. The accuracy and effectiveness of the TVM method used in the TGS image reconstruction are verified in the work, and moreover, it can save the scanning time and enhance the TGS image resolution through sparse projection sampling.
computer science, information systems,telecommunications,engineering, electrical & electronic