ABSTRACT The variety of modern X-ray sources opens possibilities to conduct measurements of a wide range of different specimens. The range includes objects with various properties such as the density, the amount and types of defects, and the size which can significantly change from sample to sample. Due to the development of portable high energy X-ray sources like betatrons, it is now possible to investigate objects of big sizes. Although, the fact that the dimensions of such objects can exceed the dimensions of X-ray setups requires to develop new types of scanning geometries and to adopt the algorithms of reconstruction and visualization. The scanning of an object along its longest dimension is one of the ways for investigating such objects. In this article, we present the step-shift scanning approach and the adaptation of filtered back-projection algorithm for this type of measurement. The developed algorithm allows to obtain the visualization of the whole volume of the object from the combination of all scanning steps or to visualize the separate parts of the object from the truncated dataset.

中文翻译：

---

用于样品步进扫描的优化滤波反投影断层扫描重建算法

摘要 现代 X 射线源的多样性开启了对各种不同样品进行测量的可能性。该范围包括具有各种特性的物体，例如密度、缺陷的数量和类型以及可能因样品而显着变化的尺寸。由于电子感应加速器等便携式高能 X 射线源的发展，现在可以研究大尺寸的物体。虽然，此类物体的尺寸可能超过 X 射线装置的尺寸这一事实需要开发新型扫描几何形状并采用重建和可视化算法。沿物体最长维度扫描物体是研究此类物体的方法之一。在本文中，我们介绍了这种类型的测量的步进扫描方法和滤波反投影算法的适应性。开发的算法允许从所有扫描步骤的组合中获得对象整个体积的可视化，或者从截断的数据集中可视化对象的单独部分。