【摘要】 讨论了卷积反投影算法实现CT图像重建时 ,被测工件旋转中心偏离理想重建中心的工程问题。基于理想卷积反投影重建算法理论 ,推导出相应校正偏离的卷积反投影算法 ,并引入单象素工件模型估计偏离参数 ;根据偏离幅度给出了理想重建算法适应的范围 ,仿真结果证明了校正算法的有效 ,利用校正算法消除了由旋转中心偏离造成的伪影 ,提高了图像分辨率更多还原

【Abstract】 The problem of displacement of center of rotation is discussed in CT image reconstruction by convolution back projection. Basing on ideal convolution back projection, a convolution back projection with a shift center of rotation has been derived for correcting images that acquired with a divergence center of rotation using the fan beam geometry with an angle equaled detector. In order to solve parameter of divergence axis of rotation, a single pixel phantom has been introduced. Software simulations are performed to certify correcting method’s feasibility, as a result, eliminate the tail off artifacts and improve the resolution of CT image.更多还原