基于重投影的多项式拟合校正射束硬化

在X射线工业CT(ICT)中,射束硬化会导致重建的图像出现伪影,甚至产生变形。为了消除这种影响,提出了一种基于重投影的多项式拟合校正射束硬化的方法。该方法对原始CT图像进行阈值分割二值化,将物体目标区域的像素值设为1;重投影此二值图像以获取X射线贯穿物体的长度集合;利用多项式拟合此长度集合与多色投影间的关系来建立射束硬化校正模型,用该模型对多色投影进行校正。与传统的多项式拟合校正方法相比,该方法不需要楔状模体(用于测量不同厚度下的衰减值,以此来建立射束硬化校正模型)。研究表明,该方法能有效地抑制射束硬化的影响。

Beam hardening correction using polynomial fitting based on reprojection

In X_ray industrial computed tomography(ICT), beam hardening caused reconstruction image artifact and distortion. To overcome this affect, a beam hardening correction method using polynomial fitting based on reprojection was presented. The method segmented the original CT image into different sections and set pixel value of the object ones to 1 to yield a binary image, then reprojected the binary image to obtain the thickness set of X_ray penetrating specimen, fitted the relationship between the thickness and the polychromatic projections by polynomial to establish the beam hardening correction model to calibrate the projections. Compared with conventional polynomial fitting correction technique, the proposed method doesn't need wedge_shaped phantom that is used to acquire the attenuation values at different thickness to establish the correction model. Research shows this method can eliminate beam hardening affect effectively.