Image calibration based on dynamic no-load data for the (co)-c-60 gantry-movable dual-projection radiography inspection system

PROCEEDINGS OF THE 2020 INTERNATIONAL CONFERENCE ON NUCLEAR ENGINEERING (ICONE2020), VOL 2(2020)

引用 0|浏览2
暂无评分
摘要
The Co-60 gantry-movable dual-projection radiography inspection system is newly developed, aiming to the inspection of vehicles entering and exiting the nuclear facilities. It adopts two Co-60 radioactive sources and two arrays of gas ionization chambers corresponding to the two sources, respectively. They can move synchronously with the gantry driven by the mechanical and control subsystem. So, dual projections could be obtained through one scan from two different directions. Compared to a single projection, the dual projections make it easier to found hidden objects and distinguish whether a dark area is due to overlapping objects or because there are well-shielded prohibited items such as nuclear materials. Therefore, it is helpful to found well-shielded nuclear materials and prevent them from being stolen. However, problems also come due to the using of two radioactive sources and the moving gantry. For the former one, it will bring about scattering effect between two sources, while for the latter one, the signals of the detectors would fluctuate as the gantry moves, owing to the vibration of collimator and gantry as well as the non-synchronous movement of sources and detectors. So, the radiography projections are needed to be corrected. In response to the second question, the no-load data is repeatedly measured when the gantry is in different positions, then a method of image calibration based on the dynamic no-load data is proposed for this inspection system to replace the correction with average no-load data. Result shows that the corrected no-load image turns smoother, meaning that the method of dynamic correction could effectively improve the radiation image.
更多
查看译文
关键词
Nuclear security, Co-60, Gantry-Movable, Image calibration
AI 理解论文
溯源树
样例
生成溯源树,研究论文发展脉络
Chat Paper
正在生成论文摘要