Abstract

Noise minimization and suppression is one of the most important problems in image processing. Background correction and detection of weak reflections were performed using the well-known algorithms implemented in the ImageJ and Digital Micrograpph (GATAN) programs and developed in this paper for experimental electron diffraction patterns obtained from a nanocrystalline 10-nm-thick hafnium oxide layer. An algorithm for processing electron diffraction patterns is proposed, which is based on a multiscale Retinex filter and Hough transform. It is also devised to present electron diffraction patterns in polar coordinates, which is often more visual and convenient for indexing reflections.

中文翻译：

---

多晶材料电子衍射图的对比度增强和背景校正方法

摘要

噪声最小化和抑制是图像处理中最重要的问题之一。背景校正和弱反射检测是使用在 ImageJ 和数字显微图 (GATAN) 程序中实现的众所周知的算法进行的，并在本文中开发，用于从 10 纳米厚的纳米晶氧化铪层获得的实验电子衍射图案。提出了一种处理电子衍射图案的算法，该算法基于多尺度 Retinex 滤波器和霍夫变换。它还被设计为在极坐标中呈现电子衍射图案，这通常更直观且便于索引反射。