摘要
针对X射线计算机断层成像(computed tomography,CT)图像重建FDK算法中,采用通常的滤波函数会导致明显的Gibbs现象,影响重建图像的质量.基于混合滤波和加权平均理论,设计了一种新型的NEW-MS-L滤波函数.先将S-L滤波函数加权平均为M3S-L滤波函数,再与NEW滤波函数叠加,得到NEW-MS-L滤波函数.通过数值仿真,分别对比NEW、R-L-S-L、R-L-NEW、R-L-MS-L和NEW-MS-L滤波函数的重建图像效果,结果表明,NEW-MS-L滤波函数能够在保持较高图像分辨率的情况下,更有效地抑制Gibbs现象,使重建图像效果更佳.
The conventional filter function of the FDK algorithm in X-ray computed tomography （CT） can result in the obvious Gibbs phenomenon, which has great effect on the CT image. A NEW-MS-L filter function is deduced based on the rationale of mixed filter theory and weighted average theory. First, the S-L filter function is weighted and averaged to form M3S-L filter function. Then, as a superposition of the M3S-L filter function and the NEW filter function, the NEW-MS-L filter function is obtained. Compared with the other filter functions such as NEW, R-L-S-L, R-L-NEW, and R-L-MS-L, the NEW-MS-L can suppress the Gibbs phenomenon while maintaining high resolution. The results show that the NEW-MS-L filter function can provide X-ray CT with better reconstruction images.
出处
《深圳大学学报（理工版）》
EI
CAS
CSCD
北大核心
2017年第3期284-289,共6页
Journal of Shenzhen University(Science and Engineering)
基金
国家自然科学基金资助项目(11674232
11074172)~~
关键词
光学工程
计算机断层成像
FDK算法
滤波函数
Gibbs现象
混合滤波
加权平均
optical engineering
computed tomography
FDK algorithm
filter function
Gibbs phenomenon
mixedfilter theorem
weighted average theorem
