| In this paper, the author discusses in detail an efficient iterative image reconstruction algorithm in industrial computerized tomography(ICT) with the narrow fan-beam projection. In order to improve the quality of reconstruction image and solve the problem that "salt and pepper" phenomena occurs in the reconstruction image, the paper proposes a novel Algorithm Reconstruction Technique(ARTâ…¡) ,compare the theory and capability of between them. The paper focuses on the ARTâ…¡'s capability when it comes to reconstruction with incomplete and noisy projection data. Iterative reconstruction method is an important kind of reconstruction methods, which has superior advantages when reconstructing images with incomplete and noisy projection data. However, in processing the projection data of inspected workpieces with large range of density, there will be "salt and pepper" in the reconstruction image of ARTâ… . Otherwise, in order to get the high quality reconstruction images ARTâ… need additional iterative times, which leads into decreasing of iterative efficiency. For the sake of solving this problem, first the paper analyzes the process of iterative method and presents the geometry and physics property of it. Considering this, the author propose a novel iterative algorithm(ARTâ…¡). The new method ameliorates the process of old iterative method and add a new iterative parameter: convergence factor, which can improve the quality of reconstruction image by changing the iterative process and adjusting convergence factor.The author completes the entire procedure from data acquisition to image reconstruction on computer. The efficiency of ARTâ…¡is evaluated in contrast to that of its traditional serial counterpart. Experiment results clearly demonstrate that ARTâ…¡can efficiently improve quality of reconstructed images when processing the incomplete projection data or noisy projection data.
|
PDF Full Text Request