Advanced tomographic reconstruction

断层扫描允许用户获得材料内部结构的三维（3D）图片，可能是在应力或温度下变形。断层成像通过拍摄对象的多个二维（2D）X射线透射图像来工作;每个图像直接类似于传统的医学X射线，但在这种情况下具有高得多的空间分辨率。物体旋转一小段距离（小于一度），每个位置都有一个图像。然后使用计算算法通过组合所有2D图像来产生结构的3D模型。该提议是为了获得能够实现这一步骤的新方法的软件和硬件，从多个2D图像到3D对象。这种新方法将特别适合于处理具有噪声图像的情况，或者对象的完整360度旋转是不可能的。

Tomography allow users to obtain three-dimensional (3D) pictures of the internal structure of a material, possibly while it is being deformed under a stress or temperature. Tomographic imaging works by taking multiple two-dimensional (2D) x-ray transmission images of an object; each image directly analogous to a conventional medical x-ray but in this case with a much higher spatial resolution. The object is rotated a small amount (less than a degree) with an image collected at each position. A computational algorithm is then used to produce a 3D model of the structure by combining all the 2D images. This proposal is to obtain the software and hardware capable of a new approach for this step, going from multiple 2D images to the 3D object. This new approach will be particularly suited to dealing with situations with noisy images, or where a full 360 degree rotation of the object is not possible.