Electrical Tomography (ET) is an advanced visualization technique, which can reconstruct all targets in an investigated field based on boundary measurements. Since the spatial resolution in the ET process can be greatly affected by the selected similarity norm, different norms may result in different ET time and spatial resolutions. In the tomographic applications nowadays, Bregman divergence (BD) has attracted increasing attention. BDs are a family of generalized similarity norm, and they can measure the similarity/difference between any two targets more accurately. Specifically, the mostly used similarity norm in the ET process (e.g., L2-norm) is only a special case of the BD family. As the key step of applying BD to the ET process, an execution method is proposed in this paper, together with the selection criteria for the optimal norm in the BD family. Simulations and experiments were conducted, and the results show that the use of an optimal BD can effectively improve the spatial resolution of an ET image.

中文翻译：

---

基于Bregman散度的电断层扫描最优相似范数

电子断层扫描 (ET) 是一种先进的可视化技术，它可以基于边界测量重建调查领域中的所有目标。由于 ET 过程中的空间分辨率受所选相似范数的影响很大，不同的范数可能会导致不同的 ET 时间和空间分辨率。在当今的层析成像应用中，布雷格曼散度（BD）越来越受到关注。BDs 是一类广义相似性范数，它们可以更准确地衡量任意两个目标之间的相似性/差异性。具体来说，ET过程中最常用的相似性范数（例如，L2-范数）只是BD家族的一个特例。作为将BD应用于ET过程的关键步骤，本文提出了一种执行方法，以及 BD 家族中最优范数的选择标准。进行了仿真和实验，结果表明使用最优BD可以有效提高ET图像的空间分辨率。