Local Kikuchi band detection in electron backscatter diffraction patterns for enhanced pattern indexing

Yitian Shen,Yongsheng Zhang,Wei Li,Hong Miao,Yongzhe Wang,Yi Zeng
DOI: https://doi.org/10.1111/jmi.13061
Abstract:Partially missing bands in the preset region for Hough transform might strongly affect the accuracy of indexing Kikuchi patterns and reduce the quality of electron backscatter diffraction (EBSD) maps. This paper proposes a novel local band detection method for such kind of low-quality patterns. The approach involves rotating bands to vertical direction, detecting the local line segments through calculating the largest horizontal average grey gradient with a constant interval of 400 pixels in vertical direction, and applying Hough transform as well as weighted averaging to these line clusters to unify the edges of Kikuchi band. Therefore, even if only part of a Kikuchi band is visible, the entire band can also be accurately extracted. The average interplanar angle error obtained by the proposed method is approximately 29.0% less than those obtained by Hough transform-based technique. Moreover, the comparison of mean angular deviation (MAD) is also discussed. The average MAD of this method is about 38.5% lower than that of Hough transform-based technique. Consequently, the local Kikuchi band detection method is expected to be used for post-processing and re-indexing the EBSD low-quality patterns.
What problem does this paper attempt to address?