Misalignment detection in X-ray microtomographic imaging system
MetadataShow full item record
X-ray microtomography becomes a promising and effective tool in various fields of internal structure studies on micrometer scale. Its resolution has been developed to reach a level far beyond the ability of a conventional medical tomography. Behind successful progressions, the quality of reconstruction images in x-ray microtomography depends on correct parameters of the system. A system calibration is a required procedure in every scanning in order to derive system parameters for reconstruction process. Without accurate values of a parameter set, a reconstruction image may be distorted or severely unrecognizable. Although being equipped with precision assistance devices, a microtomography system has a limitation by its system mechanical precision. This dissertation proposes a potential method to identify a parameter causing a misalignment effect revealing on a reconstruction result. Instead of resolving parameters in the spatial domain, major misaligned situations of metal balls are extensively analyzed in the frequency domain. The simulation study performs with a metal-sphere marker and each misaligned parameter may show its unique pattern reflecting in both power spectrum and phase of the frequency response, so a pattern defines a value of a misaligned parameter. Once knowing an incorrect variable and its value, the reconstruction process is repeated to generate a new result with better accuracy. The simulation result of this dissertation shows a possibility to detect a misalignment of x-ray microtomography system and applying this idea can enhance its reconstruction images.