欢迎访问《图学学报》 分享到：
图学学报

图学学报

• 医学图像与设备 • 上一篇    下一篇

基于正弦图的投影旋转中心校正方法

  

  1. 1. 天津医科大学生物医学工程与技术学院，天津 300070；2. 天津医科大学口腔医院，天津 300070；
    3. 天津职业技术师范大学理学院，天津 300222
  • 出版日期:2017-08-31 发布日期:2017-08-10
  • 基金资助:
    国家自然科学基金项目(81371549，81671683)；天津市自然科学基金项目(16JCYBJC28600)；天津市卫计委科技基金项目(2015KZ111)

Correction Method of the Projection Images’ Rotational Center Based on Sinogram

  1. 1. College of Biomedical Engineering, Tianjin Medical University, Tianjin 300070, China; 2. The Dental Hospital of Tianjin Medical University,
    Tianjin 300070, China; 3. The School of Science, Tianjin University of Technology and Education, Tianjin 300222, China
  • Online:2017-08-31 Published:2017-08-10

PDF (PC)

331

被引次数

1

摘要： ：CT 转台受仪器精度等影响会导致投影旋转中心偏移，这将严重影响重建图像的质
量。因此，在CT 重建前，投影旋转中心的精确校正是必不可少的。根据radon 变换推导了平
行束扫描模型投影旋转中心的计算公式，确定了正弦图中0°和180°的样本投影地址均值的一半
就是投影旋转中心的地址。根据上述方法，将sino 图中0°和180°对应的数据各复制N 次得到
图像矩阵A，同时采用“Chan-Vese”对其进行二值化分割，得到0°和180°样本投影地址并计算
投影旋转中心，最后以此为中心对sino 图在水平方向上裁剪进行校正。通过仿真和真实实验，
表明该方法能够高精度地校正投影旋转中心，有效地提高重建图像质量。

关键词: CT 重建, 正弦图, 投影旋转中心, 二值化分割, 校正

Abstract: Affected by the precision of instruments and other factors, CT turntable may cause the
deviation of projection center of rotation (COR), which will seriously affect the quality of the
reconstructed images. Thus, the precise correction of COR in projection images plays an important
role in the CT reconstruction. In this paper, a method of measuring the address of COR is proposed
based on the radon transform, which is suitable for the parallel beam scanning model. The method can
be summarized as follows: the COR address of the projection is obtained using one half of the mean
of the 0° and 180° projections’ addresses in sinogram. According to the method, the data, which
represent 0° and 180° projections in sinogram, were respectively copied by N times to get the image
matrix A. The image matrix A was then segmented into binarization with the “Chan-Vese” method,
and the 0° and 180° projections’ addresses of sample were obtained to calculate the COR address of the
projection. Finally, the COR value was regarded as the center of the sinogram, and the sinogram was
clipped properly in the horizontal direction. The simulation and experiment show that this method can
accurately correct the COR, which effectively improves the quality of the reconstructed images. method
can accurately correct the COR, which effectively improves the quality of the reconstructed images.

Key words: CT reconstruction, sinogram, center of rotation, binary segmentation, correction