Font Size: a A A

Research On The Reconstruction Techniques For Tomographic Gamma Scanning

Posted on:2004-01-13Degree:DoctorType:Dissertation
Country:ChinaCandidate:Q H ZhangFull Text:PDF
GTID:1102360095961998Subject:Nuclear Physics
Abstract/Summary:PDF Full Text Request
Tomographic gamma scanning (TGS) method is one of the most advanced non-destructive assay (NDA) methods. TGS method can determine quantitatively with high accuracy transuranic nuclides in heterogeneously distributed media with medium- and high-density, and is thus widely used to assay the location and quantity of selected radioisotopes in scraps and wastes within sealed containers.In an attempt to meet with the NDA requirements in the means of safeguards and nuclear waste management, Los Alamos National Lab (LANL) has started the work on developing TGS since the later 1980s with funding from Department of Energy (DOE) Office of Safeguards and Security. So far, TGS is already a matured technology abroad with various types of TGS setups for determining radioisotopes in the samples with medium- and high-density heterogeneous drummed scraps and wastes. However, TGS is a high-tech integration which involves in nuclear physics, nuclear radiation measurement methods, radiation image, image process, mechanical design, automatic control, computer technique, etc ... So, the TGS research has been the hot topic over the past decades. The present work is the first in China to research TGS according to the requirements of nuclear safeguards technology in China. Three key techniques, determination of detection efficiency, TGS transmission image reconstruction methods and fast reconstruction technique of TGS emission image, are researched thoroughly based on the careful analysis of lots of information on TGS.Like conventional gamma-ray spectrometry measurement method, the calibration of detection efficiency in TGS is an essential important work, the accuracy of the calibrated detection efficiency directly affects the accuracy of measurement results. However, the calibration of detection efficiency in TGS is more difficult and complicated than that in the conventional method. There are 107 detection efficiency matrix elements for a 10×10×16 (1600 voxels) TGS sample, for such a large matrix, it seems impossible and unrealistic tocalibrate experimentally as usually done for routine gamma-ray spectrometry measurement.In this work, the calibration of detection efficiency of TGS was studied using Monte Carlo method, which is very effective for obtaining the TGS detection efficiency. For a 3×3×3 TGS model, all 972 efficiency matrix elements were calculated by Monte Carlo code MCNP4B. The detection efficiency was also obtained experimentally using radioactive source 60Co. It was shown that the calculated values fitted the experimental data very well with the relative deviation less than 5%. The results show that: Monte Carlo method used to obtain TGS efficiency is feasible and reliable, and is the only practical method of determining detection efficiency in TGS.While similar to CT in principle, TGS transmission measurement has its own characteristics and requirements. CT method uses array linear detectors to perform high-resolution tomography and yield relative image of linear attenuation coefficient. TGS method utilizes a high-purity germanium (HPGe) detector with large volume to perform low-resolution tomographic transmission scans on each layer and yield absolute images of the matrix'density. So image reconstruction techniques adopted in CT could not be applied directly in TGS transmission. In this work the artificial neural network (ANN) and particle transport theory were studied, considering the TGS characteristics and the two new reconstruction techniques, ANN and Monte Carlo statistical iteration TGS transmission image reconstruction method, were proposed.In the ANN reconstruction method, the TGS transmission image reconstruction was regarded as image reorganization or parameter evaluation. The ANN principle was applied to TGS transmission image reconstruction, and then the ANN algorithm was performed in computer simulation with a TGS transmission model of 9 voxels (3x3), each voxel of 5cmx5cmx5cm. Using MATLAB software as a research platform, training samples were obtained with Monte Car...
Keywords/Search Tags:Nuclear Safeguards, Tomographic Gamma Scanning, Non-destructive assay, Monte Carlo method, Image reconstruction, Neural network
PDF Full Text Request
Related items