The Physical Design And Image Reconstruction Of X-ray CT

Industrial CT technology has the characteristics of direct, precise, nondestructive, etc. Its principle is that we can get main faults projection value through scanning the workpieces and then reconstruction fault images through a certain images reconstructed algorithm. Industrial CT can detect gray image of objects fault under nondestructive detection fault condition, with using its grayscale value to distinguish detected the internal defects, geometrical structure, material, etc. Theoretically, industrial CT testing is not affected by the materials, shape and surface conditions of the tested object. It also can precisely give object detection of two-dimensional and three-dimensional images, which has been recognized the best means of nondestructive examination now in the internationality.Now, medical CT technology and fully commercialization. While industrial CT technology have a lot of problems in industrial applications though it have gotten some progress at home and abroad since the 1980s. In order to develope and apply of the industrial CT technology further, we have studied some aspects of the physical design of X-ray industrial CT based on the works of the prior. The main contents include the structure of industrial CT system,the design and selection of X-ray source parameters, the design of mechanical scanning system and selection of performance parameters, the selection of detector probe system type, the detector array and scanning mode,the design of collimator, the calculation and analysis of CT system performance parameters such as spatial resolution, density imaging resolution and imaging time, ect. After we analyse and design the unity as well as all parts of industrial CT, we have decided to design the industry CT system with 250keV energy X-ray machine.At present, practical industrial CT system usually adopts filtered backprojection (FBP) algorithm which asked fully projection data to reconstruct the image. Because of some objective factors, we usually can not get full projection data in reality, so this method is not used. While the ART algorithm solve that perfectly by translating imaging reconstruction problem into solving linear equations. When the projection data is not full, we can see it as lacks some equations. If there is solution for the equations, we can calculate it by using the approach of iterative algorithm. So this approach is appropriate for imaging reconstruction, when we do not get full projection data. In this paper, we studied ART algorithm as well as its weight factor on the former work. In this content, weight factor was studied in order to find an optimal weight factor algorithm. We simulated the algorithm factor through C language programming in the VC platform. The experimental results have shown that the compute time of weight factor of siddon-improve algorithm is equal to that of fast algorithm. But reconstruction time of siddon-improve algorithm is half that of the fast algorithm while their image quality are quite the same. Therefore, the efficiency of reconstruction of siddon-improve algorithm is double that of fast algorithm.