论文部分内容阅读
为了加快工业CT系统的扫描速度,减小数据量,针对扇形束超短扫描问题,提出了一种带参数的基于Hilbert变换的感兴趣区域重建算法,并且在分析扇形束扫描的数据冗余基础上设计了一种适用于超短扫描的窗函数。利用Shepp-Logan头模型通过计算机模拟实验,给出了该算法和标准卷积反投影算法、F.Noo和H.Kudo算法的对比结果。实验结果表明,该方法能够准确地实现物体感兴趣区重建,并在噪声抑制能力上优于其他3种算法,因此在工业CT中具有重要的工程意义。
In order to speed up the scanning speed of industrial CT system and reduce the amount of data, a region-of-interest reconstruction algorithm with parameters based on Hilbert transform is proposed for the ultra-short scan of fan-beam. In addition, On the design of a window function suitable for ultra-short scan. Using the Shepp-Logan head model through computer simulation experiments, the algorithm and the standard convolution backprojection algorithm, the comparison results of F.Noo and H.Kudo algorithm are given. Experimental results show that the proposed method can accurately reconstruct the region of interest of interest and is superior to the other three algorithms in noise suppression. Therefore, it has important engineering significance in industrial CT.