Electron Diffraction Pattern Indexing Using Proximal Template Matching
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for indexing electron diffraction patterns, such as Hough indexing and pattern matching, face challenges with non-indexed or low-confidence locations, especially in deformed materials or nanocrystalline structures, leading to inefficiencies and potential artefacts due to reliance on correct identification of Kikuchi bands and high computational requirements.
Innovation Solution
A method that uses pattern matching combined with nearby measurement data to generate simulated templates for refining phase and orientation, reducing the number of orientations to consider and avoiding data duplication, by selecting a proximal region's phase and orientation as a starting point for simulation and refinement.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If pattern matching methods are used to improve indexing results for poor quality diffraction patterns, then indexing accuracy is improved, but analysis time increases significantly due to comparing each experimental pattern to a large number of simulated templates
Solution Approach 1:
The patent applies local quality by using the phase and orientation information from proximal (neighboring) locations to generate a limited set of candidate templates specifically relevant to the current location. Instead of comparing against all possible templates globally, the method locally adapts the template set based on the crystallographic context of surrounding areas, thereby improving indexing accuracy for poor quality patterns while reducing the computational burden.
Solution Approach 2:
The patent employs preliminary action by pre-generating a limited set of candidate templates based on the phase and orientation data from proximal locations before performing the actual pattern matching. This preliminary step filters out irrelevant templates in advance, so that when indexing is performed on poor quality patterns, only the most likely candidates are considered, thus improving accuracy without the need to exhaustively search all possible orientations.
2Reliability
If brute force pattern matching is used to index all possible orientations, then comprehensive indexing is achieved, but computational resources and time are prohibitively large
Solution Approach 1:
The patent replaces the global brute force approach with a local quality strategy by utilizing the crystallographic information from proximal locations to define a restricted set of candidate phases and orientations. This local adaptation ensures that the indexing process focuses only on relevant candidates in the vicinity of the current measurement point, maintaining reliability for comprehensive indexing while dramatically improving productivity by avoiding unnecessary comparisons with distant or irrelevant templates.
3Productivity
If Hough indexing is used for fast processing, then analysis speed is improved, but indexing fails for poor quality patterns with insufficient Kikuchi bands
Solution Approach 1:
The patent employs copying by using the phase and orientation data from successfully indexed proximal locations as a starting point for generating candidate templates at the current location. This copying of information from neighboring areas provides a robust foundation for pattern matching, enabling the method to maintain high analysis speed like Hough indexing while significantly improving the indexing success rate for poor quality patterns that lack sufficient Kikuchi bands.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
This approach significantly reduces analysis time while maintaining accuracy, avoiding data duplication errors and improving indexing quality for non-indexed or low-confidence locations, making it suitable for various applications by focusing on nearby phases and orientations, thus enhancing the efficiency and reliability of electron diffraction data indexing.
Implementation Method 1
an electron beam is focused onto a point on the surface of a sample. An electron detector is used to collect an image of the resulting diffraction pattern from that point
Implementation Method 2
the relative positions and intensities of the Kikuchi bands are dependent upon the crystal structure, composition and the 3-dimensional orientation of the crystal lattice
Data Source
AI summary
A method is provided of indexing an electron diffraction pattern obtained from a crystalline sample. Indexing data comprising phase and crystallographic orientation information is obtained for first set of locations on the sample. A second set of locations to be indexed is identified. For each nominal location in the second set an experimental electron diffraction pattern is obtained, together with a simulated template from a number of previously indexed locations in the first set, the previously indexed locations being in a proximal region of the sample to the nominal location. Further simulated templates are generated by modifying the crystallographic orientation for the previously indexed locations at angular sub-intervals. The templates are compared with the experimental pattern for the nominal location and, using a similarity measure, a resultant indexing of the location is produced. A corresponding system is also disclosed.