As a relatively new radiation imaging method, the cosmic-ray muon scattering imaging technology can be used to prevent nuclear smuggling and is of considerable significance to nuclear safety. Proposed in this paper is a new reconstruction algorithm based on density clustering, aiming to improve inspection quality with better performance. Firstly, this new algorithm is introduced in detail. Then in order to eliminate the inequity of the density threshold caused by the heterogeneity of the muon flux in different positions, a new flux correction method is proposed. Finally, three groups of simulation experiments are carried out with the help of Geant4 toolkit to optimize the algorithm parameters, verify the correction method and test the inspection quality under shielded condition, and compare this algorithm with another common inspection algorithm under different conditions. The results show that this algorithm can effectively identify and locate nuclear material with low misjudging and missing rates even when there is shielding and momentum precision is low, and the threshold correcting method is universally effective for density clustering algorithms.

宇宙射线μ子散射成像技术作为一种较新的辐射成像方法，可用于防止核走私，对核安全具有重要意义。本文提出了一种新的基于密度聚类的重建算法，旨在以更好的性能提高检测质量。首先，详细介绍了这种新算法。然后为了消除μ子通量在不同位置的不均匀性引起的密度阈值的不公平，提出了一种新的通量校正方法。最后借助Geant4工具包进行三组仿真实验，对算法参数进行优化，验证修正方法，测试屏蔽条件下的检测质量，并在不同条件下将该算法与另一种常见的检测算法进行比较。结果表明，该算法即使在存在屏蔽和动量精度较低的情况下，也能有效识别和定位核材料，误判率和遗漏率低，阈值校正方法对密度聚类算法普遍有效。