• Journal of the Korea Convergence Society
• /
• v.11 no.6
• /
• pp.1-6
• /
• 2020

In this study, research on character recognition, which is one of the fields of computer vision, was conducted. Optical character recognition, which is one of the most widely used character recognition techniques, suffers from decreasing recognition rate if the recognition target deviates from a certain standard and format. Hence, this study aimed to address this limitation by applying deep learning techniques to character recognition. In addition, as most character recognition studies have been limited to English or number recognition, the recognition range has been expanded through additional data training on Korean text. As a result, this study derived a deep learning-based character recognition algorithm for Korean text recognition. The algorithm obtained a score of 0.841 on the 1-NED evaluation method, which is a similar result to that of English recognition. Further, based on the analysis of the results, major issues with Korean text recognition and possible future study tasks are introduced.

• Proceedings of the Korea Contents Association Conference
• /
• 2009.05a
• /
• pp.927-932
• /
• 2009

There is such a kind of characteristic by using the computer digitization technology's interaction. It not only provides the authoring tool, and also has the user-friendly window. The paper based on this kind of interaction, through the sound and the color exchange of information, as well as each information's digitization, carries on about the scientific dielectric related advance research. This research carries on the analysis to the people to the color and the sound perceptual characteristic first, then makes comparison to the people regarding the color and the music perceptual cognition.. The research is about the research stage of the interactive sound installation creation, which shape is carried on the analysis through the extraction image in color. It is based on the perceptual characteristic,is that the algorithm manufacture reasearch of how to make color sonification. Through capturing the non-stop changing colors by camera,it transforms colors into music. The paper's foundation is the fundamental algorithm manufacture of experimental music composition.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.16 no.10
• /
• pp.2121-2128
• /
• 2012

This paper proposed the real time face area detection using Wavelet transform and the strong detection algorithm that satisfies the efficiency of computation and detection performance at the same time was proposed. The detected face image recognizes the face by configuring the low-dimensional face symbol through the principal component analysis. The proposed method is well suited for real-time system construction because it doesn't require a lot of computation compared to the existing geometric feature-based method or appearance-based method and it can maintain high recognition rate using the minimum amount of information. In addition, in order to reduce the wrong recognition or recognition error occurred during face recognition, the input symbol of Hidden Markov Model is used by configuring the feature values projected to the unique space as a certain symbol through clustering algorithm. By doing so, any input face will be recognized as a face model that has the highest probability. As a result of experiment, when comparing the existing method Euclidean and Mahananobis, the proposed method showed superior recognition performance in incorrect matching or matching error.

• Journal of KIISE:Software and Applications
• /
• v.27 no.8
• /
• pp.815-826
• /
• 2000

This study is for the segmentation and volume calculation of the white matter and gray matter from brain MRI. In general, the volume of white and gray matter is reduced by contraction of each components in the case of mental retardation which are Alzheimer's disease and Down's syndrome. As results, it is useful for diagnostic and early detection for various mental retardation through the tracing of variation for its volume from the brain MRI. But, until now, it was very difficult to calculate the partial volume of each components existing in some thickness, because MR image was represented by single gray value after scanning by MR scanner. Accordingly, new segmentation algorithm proposed in this paper is to calculate the partial volume of the white and gray matter existing in some thickness through the analysis of the blurred gray value, and is to determine the threshold for segmentation of white and gray matter, and is to calculate the volume of each segmented component. And finally, proposed algorithm was applied the models which was created manually, and then acquired results was compared with that of original model.

• Imaging Science in Dentistry
• /
• v.51 no.1
• /
• pp.1-7
• /
• 2021

Purpose: The aim of this study was to assess artifacts generated in cone-beam computed tomography (CBCT) of 3 types of dental implants using 3 metal artifact reduction (MAR) algorithm conditions (pre-acquisition MAR, post-acquisition MAR, and no MAR), and 2 peak kilovoltage (kVp) settings. Materials and Methods: Titanium-zirconium, titanium, and zirconium alloy implants were placed in a dry mandible. CBCT images were acquired using 84 and 90 kVp and at normal resolution for all 3 MAR conditions. The images were analyzed using ImageJ software (National Institutes of Health, Bethesda, MD) to calculate the intensity of artifacts for each combination of material and settings. A 3-factor analysis of variance model with up to 3-way interactions was used to determine whether there was a statistically significant difference in the mean intensity of artifacts associated with each factor. Results: The analysis of all 3 MAR conditions showed that using no MAR resulted in substantially more severe artifacts than either of the 2 MAR algorithms for the 3 implant materials; however, there were no significant differences between pre- and post-acquisition MAR. The 90 kVp setting generated less intense artifacts on average than the 84 kVp setting. The titanium-zirconium alloy generated significantly less intense artifacts than zirconium. Titanium generated artifacts at an intermediate level relative to the other 2 implant materials, but was not statistically significantly different from either. Conclusion: This in vitro study suggests that artifacts can be minimized by using a titanium-zirconium alloy at the 90 kVp setting, with either MAR setting.

• Proceedings of the Korean Society of Computer Information Conference
• /
• 2021.07a
• /
• pp.701-702
• /
• 2021

The purpose of the study was to identify fiber changes induced by electrical stimulation of a certain neural substrate in the rat brain. In the stimulation group, the peripheral nerve was injured and the brain area associated to inhibit sensory information was electrically stimulated. There were sham and sham stimulation groups as controls. Then high-field diffusion tensor imaging (DTI) was acquired. 35 features were taken from the DTI measures from 7 different brain pathways. To compare the efficacy of the classification for 3 animal groups, the linear regression analysis (LDA) and the machine learning technique (MLP) were applied. It was found that the testing accuracy by MLP was about 77%, but that of accuracy by LDA was much higher than MLP. In conclusion, machine learning algorithm could be used to identify and predict the changes of the brain white matter in some situations. The limits of this study will be discussed.

• Korean Journal of Construction Engineering and Management
• /
• v.25 no.2
• /
• pp.36-44
• /
• 2024

This paper presents the application of Convolutional Neural Networks (CNNs) and Region of Interest (ROI) techniques for concrete crack analysis. Surfaces of concrete structures, such as beams, etc., are exposed to fatigue stress and cyclic loads, typically resulting in the initiation of cracks at a microscopic level on the structure's surface. Early detection enables preventative measures to mitigate potential damage and failures. Conventional manual inspections often yield subpar results, especially for large-scale infrastructure where access is challenging and detecting cracks can be difficult. This paper presents data collection, edge segmentation and ROI techniques application, and analysis of concrete cracks using Convolutional Neural Networks. This paper aims to achieve the following objectives: Firstly, achieving improved accuracy in crack detection using image-based technology compared to traditional manual inspection methods. Secondly, developing an algorithm that utilizes enhanced Sobel edge segmentation and ROI techniques. The algorithm provides automated crack detection capabilities for non-destructive testing.

• Journal of the Korean Society of Radiology
• /
• v.15 no.5
• /
• pp.591-602
• /
• 2021

Acute ischemic stroke(AIS) should be diagnosed within a few hours of onset of cerebral infarction symptoms using diagnostic radiology. In this study, we evaluated the clinical usefulness of SVD and the Bayesian algorithm to measure the volume of cerebral infarction using computed tomography perfusion(CTP) imaging and magnetic resonance diffusion-weighted imaging(MR DWI). We retrospectively included 50 patients (male : female = 33 : 17) who visited the emergency department with symptoms of AIS from September 2017 to September 2020. The cerebral infarct volume measured by SVD and the Bayesian algorithm was analyzed using the Wilcoxon signed rank test and expressed as a median value and an interquartile range of 25 - 75 %. The core volume measured by SVD and the Bayesian algorithm using was CTP imaging was 18.07 (7.76 - 33.98) cc and 47.3 (23.76 - 79.11) cc, respectively, while the penumbra volume was 140.24 (117.8 - 176.89) cc and 105.05 (72.52 - 141.98) cc, respectively. The mismatch ratio was 7.56 % (4.36 - 15.26 %) and 2.08 % (1.68 - 2.77 %) for SVD and the Bayesian algorithm, respectively, and all the measured values had statistically significant differences (p < 0.05). Spearman's correlation analysis showed that the correlation coefficient of the cerebral infarct volume measured by the Bayesian algorithm using CTP imaging and MR DWI was higher than that of the cerebral infarct volume measured by SVD using CTP imaging and MR DWI (r = 0.915 vs. r = 0.763 ; p < 0.01). Furthermore, the results of the Bland Altman plot analysis demonstrated that the slope of the scatter plot of the cerebral infarct volume measured by the Bayesian algorithm using CTP imaging and MR DWI was more steady than that of the cerebral infarct volume measured by SVD using CTP imaging and MR DWI (y = -0.065 vs. y = -0.749), indicating that the Bayesian algorithm was more reliable than SVD. In conclusion, the Bayesian algorithm is more accurate than SVD in measuring cerebral infarct volume. Therefore, it can be useful in clinical utility.

• Journal of Korea Multimedia Society
• /
• v.21 no.2
• /
• pp.130-137
• /
• 2018

Recently, human life is getting longer due to change of living environment and development of medical technology, and silver medical technology has been in the limelight. Geriatric skin disease is difficult to detect early, and when it is missed, it becomes a malignant disease and is difficult to treatment. Melanoma is one of the most common diseases of geriatric skin disease and initially has a similar modality with the nevus. In order to overcome this problem, we attempted to perform a feature analysis in order to attempt automatic detection of melanoma-like lesions. In this paper, one is first order analysis using information of pixels in radiomic feature. The other is a gray-level co-occurrence matrix and a gray level run length matrix, which are feature extraction methods for converting image information into a matrix. The features were extracted through these analyses. And classification is implemented by SVM.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.15 no.7
• /
• pp.1439-1446
• /
• 2011

This paper describes the algorithm that lowers the dimension, maintains the object recognition and significantly reduces the eigenspace configuration time by combining the higher correlation feature information and Principle Component Analysis. Since the suggested method doesn't require a lot of computation than the method using existing geometric information or stereo image, the fact that it is very suitable for building the real-time system has been proved through the experiment. In addition, since the existing point to point method which is a simple distance calculation has many errors, in this paper to improve recognition rate the recognition error could be reduced by using several successive input images as a unit of recognition with K-Nearest Neighbor which is the improved Class to Class method.