A new efficient imaging reconstruction method for muon scattering tomography
Pei Yu,Ziwen Pan,Zhengyang He,Li Deng,Yu Xu,Yuhong Yu,Xueheng Zhang,Zechao Kang,Zhe Chen,Zebin Lin,Liangwen Chen,Lei Yang,Zhiyu Sun
DOI: https://doi.org/10.1016/j.nima.2024.169932
2024-10-07
Abstract:Muon scattering tomography (MST) is a promising technique in nuclear security inspection, utilizing its multiple scattering property which is sensitive to elements with high atomic numbers, and is also correlated with the material density. A proper imaging reconstruction method is essential for revealing the spatial distribution of materials within an object from measurements. This work proposes a new imaging reconstruction method called "big voxel and angle capping". The scattering angle is reconstructed using the conventional Point of Closest Approach (PoCA) method. In the new method, reconstructed scattering angles belonging to voxels in an l×m×n matrix are aggregated into the center voxel, termed the "big voxel". This significantly reduces statistical fluctuations in each big voxel. Subsequently, the influence of large scattering angles on determining material densities in big voxels is mitigated using a capping angle. The proposed algorithm is tested on the experimental data comprising approximately 1.8×104 muon scattering events detected in ∼ 24 h. The mean square error (MSE) and the structural similarity index measure (SSIM) indices are employed to quantitatively assess the imaging quality and optimize the new method. After scanning the two major parameters, the optimal parameter space is achieved with the big voxel size of ∼ 9, and the capping angle of ∼4∘ for tungsten blocks. The imaging quality is more sensitive to the capping angle than the big voxel size. An appropriate capping angle effectively removes large artifacts in reconstructed images caused by muon events with large scattering angles. This method is efficient in terms of both time and storage consumption.
physics, particles & fields, nuclear,nuclear science & technology,instruments & instrumentation