Analyzer Image Processing for Crisp Elemental Maps
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In scanning electron microscopes, elemental maps with low signal-to-noise ratios require longer measurement times due to low characteristic X-ray production efficiency, leading to noise reduction methods that also blur edges, while short measurement times result in statistical errors misinterpreted as edges, weakening the filter effect.
Innovation Solution
An analyzer apparatus and method that scans a sample to produce secondary and backscattered electron images and elemental maps, using edge-preserving smoothing filters based on high signal-to-noise ratio backscattered electron images to reduce noise in elemental maps while preserving edges, by reflecting edge information from the backscattered electron images into the filtering process.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If the measurement time is extended to reduce noise in elemental maps, then the signal-to-noise ratio improves, but the productivity decreases
Solution Approach 1:
The patent applies spatial filtering (such as averaging filters or Gaussian filters) to the elemental map data before final analysis. This preliminary processing step accumulates pixel intensities and reduces statistical errors, thereby improving the signal-to-noise ratio without requiring extended measurement times. The filter operates on the already-acquired data, effectively reducing noise while maintaining the original measurement duration.
2Measurement precision
If the kernel size of the spatial filter is increased to reduce noise more effectively, then the noise reduction improves, but the edges become blurred
Solution Approach 1:
The patent employs edge-preserving smoothing filters that adaptively adjust the filtering strength based on local image characteristics. In regions with high intensity gradients (edges), the filter reduces its smoothing effect to preserve edge sharpness, while in homogeneous regions, it applies stronger smoothing to reduce noise. This local adaptation allows the use of larger kernel sizes for noise reduction without compromising edge definition.
3Manufacturing precision
If an edge preserving smoothing filter is used to maintain edges, then edge sharpness is preserved, but statistical errors are misinterpreted as edges, weakening the filter effect
Solution Approach 1:
The patent applies a preliminary spatial filtering step (such as averaging or Gaussian filtering) before applying the edge-preserving smoothing filter. This initial filtering reduces the statistical errors and intensity variations that would otherwise be misinterpreted as edges. By pre-processing the elemental map data to reduce random fluctuations, the subsequent edge-preserving filter can more accurately distinguish between true edges and statistical noise, thereby maintaining its effectiveness even in short measurement conditions.
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 allows for the production of clearer elemental maps with reduced noise while maintaining edge clarity, even with large kernel sizes, by applying filters based on high signal-to-noise ratio images to those with lower ratios, enhancing image crispness.
Implementation Method 1
The scanning of the sample with the probe emits secondary electrons, backscattered electrons, characteristic X-rays, and so on from the sample
Implementation Method 2
The scanning of the sample with the probe emits secondary electrons, backscattered electrons, characteristic X-rays, and so on from the sample
Implementation Method 3
The scanning of the sample with the probe emits secondary electrons, backscattered electrons, characteristic X-rays, and so on from the sample
Implementation Method 4
by applying a spatial filter such as an averaging filter or a Gaussian filter to the elemental map, the intensities of pixels are accumulated, whereby the statistical error can be reduced
Data Source
AI summary
There is provided an analyzer apparatus capable of generating crisp scanned images. In the analyzer apparatus, a sample is scanned with a probe such that a first signal and a second signal are emitted from the sample. The analyzer apparatus comprises: a first detector for detecting the first signal and producing a first detector signal; a second detector for detecting the second signal and producing a second detector signal; and an image processing unit operating (i) to produce a first scanned image and a second scanned image from the first detector signal and the second detector signal, respectively, (ii) to create a filter based on the second scanned image having a higher signal-to-noise ratio than that of the first scanned image, and (iii) to apply the filter to the first scanned image.