The most troublesome problem in automated X-ray mask inspection is how to exactly determine the threshold level for extracting the pattern portions of each scanning electron microscopic (SEM) image. An exact determination is difficult because the histogram shows, in most cases, a complicated multi-modal pattern and the true threshold level often varies with each successive image. This paper presents a novel thresholding approach for segmenting SEM images of X-ray masks. In this approach, the shape of the histogram of each image is iteratively analyzed until a threshold value minimizing the cost of the correspondence with a reference histogram and satisfying criteria for determining thresholds is obtained. This new approach is used in an automated inspection system. When the input image resolution is set to 0.05µm/pixel, experiments confirm 0.1µm defects are unfailingly detected.
The copyright of the original papers published on this site belongs to IEICE. Unauthorized use of the original or translated papers is prohibited. See IEICE Provisions on Copyright for details.
Copy
Minoru ITO, "An Automated Thresholding Approach for Segmenting Deteriorated SEM Images in X-Ray Mask Visual Inspection" in IEICE TRANSACTIONS on Information,
vol. E79-D, no. 6, pp. 866-872, June 1996, doi: .
Abstract: The most troublesome problem in automated X-ray mask inspection is how to exactly determine the threshold level for extracting the pattern portions of each scanning electron microscopic (SEM) image. An exact determination is difficult because the histogram shows, in most cases, a complicated multi-modal pattern and the true threshold level often varies with each successive image. This paper presents a novel thresholding approach for segmenting SEM images of X-ray masks. In this approach, the shape of the histogram of each image is iteratively analyzed until a threshold value minimizing the cost of the correspondence with a reference histogram and satisfying criteria for determining thresholds is obtained. This new approach is used in an automated inspection system. When the input image resolution is set to 0.05µm/pixel, experiments confirm 0.1µm defects are unfailingly detected.
URL: https://global.ieice.org/en_transactions/information/10.1587/e79-d_6_866/_p
Copy
@ARTICLE{e79-d_6_866,
author={Minoru ITO, },
journal={IEICE TRANSACTIONS on Information},
title={An Automated Thresholding Approach for Segmenting Deteriorated SEM Images in X-Ray Mask Visual Inspection},
year={1996},
volume={E79-D},
number={6},
pages={866-872},
abstract={The most troublesome problem in automated X-ray mask inspection is how to exactly determine the threshold level for extracting the pattern portions of each scanning electron microscopic (SEM) image. An exact determination is difficult because the histogram shows, in most cases, a complicated multi-modal pattern and the true threshold level often varies with each successive image. This paper presents a novel thresholding approach for segmenting SEM images of X-ray masks. In this approach, the shape of the histogram of each image is iteratively analyzed until a threshold value minimizing the cost of the correspondence with a reference histogram and satisfying criteria for determining thresholds is obtained. This new approach is used in an automated inspection system. When the input image resolution is set to 0.05µm/pixel, experiments confirm 0.1µm defects are unfailingly detected.},
keywords={},
doi={},
ISSN={},
month={June},}
Copy
TY - JOUR
TI - An Automated Thresholding Approach for Segmenting Deteriorated SEM Images in X-Ray Mask Visual Inspection
T2 - IEICE TRANSACTIONS on Information
SP - 866
EP - 872
AU - Minoru ITO
PY - 1996
DO -
JO - IEICE TRANSACTIONS on Information
SN -
VL - E79-D
IS - 6
JA - IEICE TRANSACTIONS on Information
Y1 - June 1996
AB - The most troublesome problem in automated X-ray mask inspection is how to exactly determine the threshold level for extracting the pattern portions of each scanning electron microscopic (SEM) image. An exact determination is difficult because the histogram shows, in most cases, a complicated multi-modal pattern and the true threshold level often varies with each successive image. This paper presents a novel thresholding approach for segmenting SEM images of X-ray masks. In this approach, the shape of the histogram of each image is iteratively analyzed until a threshold value minimizing the cost of the correspondence with a reference histogram and satisfying criteria for determining thresholds is obtained. This new approach is used in an automated inspection system. When the input image resolution is set to 0.05µm/pixel, experiments confirm 0.1µm defects are unfailingly detected.
ER -