• Smart Structures and Systems
• /
• v.29 no.1
• /
• pp.29-39
• /
• 2022

Crack detection plays an important role in the maintenance and protection of steel box girder of bridges. However, since the cracks only occupy an extremely small region of the high-resolution images captured from actual conditions, the existing methods cannot deal with this kind of image effectively. To solve this problem, this paper proposed a novel three-stage method based on deep learning technology and morphology operations. The training set and test set used in this paper are composed of 360 images (4928 × 3264 pixels) in steel girder box. The first stage of the proposed model converted high-resolution images into sub-images by using patch-based method and located the region of cracks by CBAM ResNet-50 model. The Recall reaches 0.95 on the test set. The second stage of our method uses the Attention U-Net model to get the accurate geometric edges of cracks based on results in the first stage. The IoU of the segmentation model implemented in this stage attains 0.48. In the third stage of the model, we remove the wrong-predicted isolated points in the predicted results through dilate operation and outlier elimination algorithm. The IoU of test set ascends to 0.70 after this stage. Ablation experiments are conducted to optimize the parameters and further promote the accuracy of the proposed method. The result shows that: (1) the best patch size of sub-images is 1024 × 1024. (2) the CBAM ResNet-50 and the Attention U-Net achieved the best results in the first and the second stage, respectively. (3) Pre-training the model of the first two stages can improve the IoU by 2.9%. In general, our method is of great significance for crack detection.

• The Journal of Korean Medicine
• /
• v.21 no.4
• /
• pp.148-158
• /
• 2000

Objective : The purpose of this study was to investigate the back shape of school children using the phase-shifting scanning grating projection moire interferometer, which was developed by the Korea Advanced Institute of Science and Technology and is useful in evaluating three dimensional back shape. Methods : In this study the subjects consisted of 1,358 pupils [711 boys(52.36%), 647 girls(47.64%)] attending elementary and middle schools in Seoul. Their ages ranged from nine to fifteen and the average age was 12.2. With the phase-shifting scanning grating projection moire interferometer, the posterior view of the body were taken to see if there are correlations of moire fringe number, width difference between left and right, and correlation between differences in moire fringe number and width on both sides in the scapular, lumbar and gluteal regions. Results : The results were as follows : I. More frequent findings of fringe were observed on the right in all regions : in the scapular region, 309 boys(43.4%) and 156 girls(24.2%) had more fringe numbers on the right side; in the lumbar region, 68 boys(9.5%) and 11 girls(1.7%); and in the gluteal region, 160 boys(22.4%) and 63 girls(9.8%). Such tendency was striking especially in the scapular and lumbar regions, and in boys rather than in girls. In the scapula, 661 subjects(48.7%) with one moire fringe on either side need further attention and 110 subjects(8.I %) with two or more are required to do follow-up radiography for scoliosis. 2. In an analysis of width difference in the trunk, the left side is wider in all regions except for the gluteal region in boys : in the scapular region 21 boys(3.0%) and 103 girls(15.9%); in the lumbar region, 87 boys(12.2%) and 250 girls(38.6%); and in the gluteal region 197 girls(30.4%) had a wider left side and 45 boys(6.3%) showed a wider right side. 3. In correlation analysis of the number of moire fringe and width difference in each region, the side where more moire fringes were observed was significantly wider in the lumbar and gluteal regions, but not in the scapular region.(p<0.01) Conclusions : From these results, it is concluded that the back shape of elementary and middle school students in Seoul shows that the right side had more moire fringes; the left side was wider; and especially in the lumbar and gluteal regions the side where more moire fringes were observed was wider.

• The Journal of the Korean bone and joint tumor society
• /
• v.1 no.2
• /
• pp.244-248
• /
• 1995

Glomus tumor is uncommon tumor with an incidence of about 1% in soft tissue tumors and the single most common location is the subungual region of the finger. Many reports have emphasized that glomus tumors were difficult to diagnose and that the result of surgical removal was good, but nail deformity has received little attention. Most of authors removed the entire nail or partially splitted the nail to exposure the nail bed and tumor. Nobody reported whether nail deformity after operation was present or not. But, an abnormal nail is both a cosmetic and a functional problem in that catches on objects. In this study, our attention was postoperative nail deformity as well as the results of surgical removal. Seven patients(eight cases) who were diagnosed as a subungal glomus tumor in the hand were treated by surgical excision without removal of the nail and followed over twenty two months. And then we report on our experience with 8 cases of subungual glomus tumor which were successfully treated by complete excision without nail deformity.

• Journal of Computing Science and Engineering
• /
• v.10 no.4
• /
• pp.111-117
• /
• 2016

An accurate approach for diagnosis of attention deficit hyperactivity disorder (ADHD) is presented in this paper. The presented technique efficiently classifies three subtypes of ADHD (ADHD-C, ADHD-H, ADHD-I) and typically developing control (TDC) by using only structural magnetic resonance imaging (MRI). The research examines structural MRI of the hippocampus from the ADHD-200 database. Each available MRI has been processed by a region-of-interest (ROI) to build a set of features for further analysis. The presented ADHD diagnostic approach unifies feature selection and classification techniques. The feature selection technique based on the proposed binary-coded genetic algorithm searches for an optimal subset of features extracted from the hippocampus. The classification technique uses a chosen optimal subset of features for accurate classification of three subtypes of ADHD and TDC. In this study, the famous Extreme Learning Machine is used as a classification technique. Experimental results clearly indicate that the presented BCGA-ELM (binary-coded genetic algorithm coupled with Extreme Learning Machine) efficiently classifies TDC and three subtypes of ADHD and outperforms existing techniques.

• The Korean Journal of Applied Statistics
• /
• v.25 no.3
• /
• pp.389-401
• /
• 2012

This paper derives a benchmark dose(BMD) and its 95% lower confidence limit(BMDL) using a semi-parametric regression model for small lead based changes in attention-deficit hyperactivity disorder(ADHD) scores in the first wave of the Children's Health and Environment Research(CHEER) survey data, which have been regularly collected in South Korea since 2005. Ha et al. (2009) showed that the appearance of ADHD symptoms had a borderline trend of increasing with the blood lead concentration. Butdz-J${\o}$rgensen (EFSA, 2010a) derived the BMDL of lead corresponding to a benchmark region of 1 full intelligent quotient (IQ) score using the raw data in Lanphear et al. (2005, EHP). European Food Safety Authority (EFSA, 2010b) determined the BMDL of $1.2{\mu}g/dl$ as a reference point for the characterization of lead when assessing the risk of the intellectual deficit measured by IQ scores. Kim et al. (2011) indicated that an even lower BMDL could be obtained based on the ADHD score; however, the BMDLs depended heavily upon the model assumptions. We show in this paper that a semi-parametric approach resolves the model dependence of BMDLs.

• English & American cultural studies
• /
• v.18 no.1
• /
• pp.93-120
• /
• 2018

Constance Fenimore Woolson is among those whom scholars have for long been trying to rediscover and add to the list of representative American writers. The primary methodology has been regionalism, based on the fact that most of her work portrays remote, exotic regions in and out of America. Still, Woolson remains obscure to general readers as well as literary critics outside a small circle of her scholarship. This essay attributes that obscurity to Woolson scholars' blind reliance on regionalism's nationalistic assumption in reading Woolson's multifaceted writing, and proposes to explore her nationally and regionally displacing view of the rigidly stereotypical and ideologically biased binary opposition between the center and the margin in postbellum America. The essay takes as an example the only story by Woolson that has never been reprinted or anthologized until very recently, "In Sloane Street," and examines why it resists the scholarly endeavor to regional categorization. The examination especially focuses on how the story exposes the Americanizing conceptualization of the region and its limits. The essay concludes with an attention to the story's ending where Woolson abruptly yet deliberately introduces a form of almanac as the main character Gertrude's mode of daily record. The attention to that uniquely hybrid genre in the American literary tradition, which encompasses the public and the private, the universal and the local, sheds light on Woolson's authorial intention to deconstruct the Manichean view of literary regionalism.

• Journal of Information Processing Systems
• /
• v.18 no.4
• /
• pp.575-586
• /
• 2022

Uneven light in real-world causes visual degradation for underexposed regions. For these regions, insufficient consideration during enhancement procedure will result in over-/under-exposure, loss of details and color distortion. Confronting such challenges, an unsupervised low-light image enhancement network is proposed in this paper based on the guidance of the unpaired low-/normal-light images. The key components in our network include super-resolution module (SRM), a GAN-based low-light image enhancement network (LLIEN), and denoising-scaling module (DSM). The SRM improves the resolution of the low-light input images before illumination enhancement. Such design philosophy improves the effectiveness of texture details preservation by operating in high-resolution space. Subsequently, local lightness attention module in LLIEN effectively distinguishes unevenly illuminated areas and puts emphasis on low-light areas, ensuring the spatial consistency of illumination for locally underexposed images. Then, multiple discriminators, i.e., global discriminator, local region discriminator, and color discriminator performs assessment from different perspectives to avoid over-/under-exposure and color distortion, which guides the network to generate images that in line with human aesthetic perception. Finally, the DSM performs noise removal and obtains high-quality enhanced images. Both qualitative and quantitative experiments demonstrate that our approach achieves favorable results, which indicates its superior capacity on illumination and texture details restoration.

• Proceedings of the Korean Institute of Surface Engineering Conference
• /
• 2016.11a
• /
• pp.181-181
• /
• 2016

Hydrogenated microcrystalline silicon (${\mu}c-Si:H$) films have attracted much attention as materials of the bottom-cells in Si thin film tandem photovoltaics due to their low bandgap and excellent stability against light soaking. However, in PECVD, the source gas $SiH_4$ must be highly diluted by $H_2$, which eventually results in low deposition rate. Moreover, it is known that high-rate ${\mu}c-Si:H$ growth is usually accompanied by a large number of dangling-bond (DB) defects in the resulting films, which act as recombination centers for photoexcited carriers, leading to a deterioration in the device performance. During film deposition, Si nanoparticles generated in $SiH_4$ discharges can be incorporated into films, and such incorporation may have effects on film properties depending on the size, structure, and volume fraction of nanoparticles incorporated into films. Here we report experimental results on the effects of nonoparticles incorporation at the different substrate temperature studied using a multi-hollow discharge plasma CVD method in which such incorporation can be significantly suppressed in upstream region by setting the gas flow velocity high enough to drive nanoparticles toward the downstream region. All experiments were performed with the multi-hollow discharge plasma CVD reactor at RT, 100, and $250^{\circ}C$, respectively. The gas flow rate ratio of $SiH_4$ to $H_2$ was 0.997. The total gas pressure P was kept at 2 Torr. The discharge frequency and power were 60 MHz, 180 W, respectively. Crystallinity Xc of resulting films was evaluated using Raman spectra. The defect densities of the films were measured with electron spin resonance (ESR). The defect density of fims deposited in the downstream region (with nonoparticles) is higher defect density than that in the upstream region (without nanoparticles) at low substrate temperature of RT and $100^{\circ}C$. This result indicates that nanoparticle incorporation can change considerably their film properties depending on the substrate temperature.

• Imaging Science in Dentistry
• /
• v.47 no.4
• /
• pp.261-267
• /
• 2017

Purpose: Analyzing different patterns of maxillary sinus septa in cone-beam computed tomography (CBCT) images and predicting maxillary sinus membrane perforations. Materials and Methods: In this cross-sectional study, CBCT images of 222 patients ranging from 20 to 81 years old were evaluated. One hundred fifty-two patients (93 females and 59 males) who had maxillary sinus septa in axial views were included in this study. Cross-sectional images were used to determine classifications of sinus septa and the risk of membrane perforation using a method modified from Al-Faraje et al. Variables of sex, age, and dental status were considered. Chi-squared and Kruskal-Wallis tests were used for data analysis(P<.05). Results: In this study, 265 maxillary sinus septal patterns were found. The mean age of the patients was $44.1{\pm}14.7$ years old. The Class I and VII-div II patterns had the greatest and least prevalence, respectively. Furthermore, there was a significant relationship between the location of septa and the frequency of membrane perforation risk (P<.05). In this study, the relationship of different patterns of septa with dental status did not differ significantly (P>0.05). Conclusion: A higher prevalence of moderate risk of membrane perforation in the molar region relative to the premolar region was observed. Furthermore, maxillary sinus septa occur most frequently in the molar region, demonstrating the importance of paying attention to this region during sinus lift surgery. This study did not show any relationship between tooth loss and the presence of septa.

• Lingua Humanitatis
• /
• v.8
• /
• pp.265-289
• /
• 2006

The geographical characteristics of Yeongdong(永同) the southernmost part of the Chungcheongbukdo province, has attracted attention among the academic circle as one of the dialectal contact regions since it adjoins the Gyeongsang and Jeolla dialects. Unlike the local language in Mooju (Jellado dialect) adjacent to the Southwest part, the local language in Yeongdong is quite different from that of Kimcheon (Gyeongsang dialect). More specifically, it is noteworthy that the boundary line of the Gyeongsang dialect is found in this region, which is different from the administrative division. In other words, the local language in Yeongdong is divided into the Chungcheong dialect and the Gyeongsang dialect, and furthermore each dialect region still has the characteristics of the other region's dialect. For example, the phonological structure of Yeongdong Chungcheongdo dialect has very unique characteristics of the fudged dialect, which is seemingly influenced by the Gyeongsang dialect. The present study is to define the bottom boundary line of the southeast area of the Chungcheong dialect by identifying the boundary line between the Gyeongsang dialect and the Chungcheong dialect, and to clarify its specific sound system generated by the contact of these two dialects. For this, the author collected and analyzed data of the local language around Yeongdong and adjacent areas. It was found that Cheongwha-ri, Deokjin-ri, and Sanjeo-ri at Yeongsan-myeon, and Mugeunjeom, Sangga-ri, and Jungga-ri at Yeongdong-eup, among the regions that belongs to Chungcheong dialect within the local language of Yeongdong, show the characteristics of the Gyeongsang dialect. Accordingly, the western areas of these villages become the southeast boundary line of the Chungcheong dialect. Also, the unique phonological characteristics of the Yeongdong Chungcheong dialect is affected by the Gyeongsang dialect, among which "rhythms, y deletion, nasal phoneme deletion, and w deletion" appeared. It is thought to be the unique fudged dialectal phenomenon that appeared only in this region. The research result is expected to be of some help in finding out various aspects of dialectal contacts as well as clarifying the phonological features of the local language in Yeongdong, and thereby contributing to exact divisioning of the Chungcheong dialect.