• Journal of Intelligence and Information Systems
• /
• v.24 no.1
• /
• pp.1-23
• /
• 2018

From the 21st century, various high-quality services have come up with the growth of the internet or 'Information and Communication Technologies'. Especially, the scale of E-commerce industry in which Amazon and E-bay are standing out is exploding in a large way. As E-commerce grows, Customers could get what they want to buy easily while comparing various products because more products have been registered at online shopping malls. However, a problem has arisen with the growth of E-commerce. As too many products have been registered, it has become difficult for customers to search what they really need in the flood of products. When customers search for desired products with a generalized keyword, too many products have come out as a result. On the contrary, few products have been searched if customers type in details of products because concrete product-attributes have been registered rarely. In this situation, recognizing texts in images automatically with a machine can be a solution. Because bulk of product details are written in catalogs as image format, most of product information are not searched with text inputs in the current text-based searching system. It means if information in images can be converted to text format, customers can search products with product-details, which make them shop more conveniently. There are various existing OCR(Optical Character Recognition) programs which can recognize texts in images. But existing OCR programs are hard to be applied to catalog because they have problems in recognizing texts in certain circumstances, like texts are not big enough or fonts are not consistent. Therefore, this research suggests the way to recognize keywords in catalog with the Deep Learning algorithm which is state of the art in image-recognition area from 2010s. Single Shot Multibox Detector(SSD), which is a credited model for object-detection performance, can be used with structures re-designed to take into account the difference of text from object. But there is an issue that SSD model needs a lot of labeled-train data to be trained, because of the characteristic of deep learning algorithms, that it should be trained by supervised-learning. To collect data, we can try labelling location and classification information to texts in catalog manually. But if data are collected manually, many problems would come up. Some keywords would be missed because human can make mistakes while labelling train data. And it becomes too time-consuming to collect train data considering the scale of data needed or costly if a lot of workers are hired to shorten the time. Furthermore, if some specific keywords are needed to be trained, searching images that have the words would be difficult, as well. To solve the data issue, this research developed a program which create train data automatically. This program can make images which have various keywords and pictures like catalog and save location-information of keywords at the same time. With this program, not only data can be collected efficiently, but also the performance of SSD model becomes better. The SSD model recorded 81.99% of recognition rate with 20,000 data created by the program. Moreover, this research had an efficiency test of SSD model according to data differences to analyze what feature of data exert influence upon the performance of recognizing texts in images. As a result, it is figured out that the number of labeled keywords, the addition of overlapped keyword label, the existence of keywords that is not labeled, the spaces among keywords and the differences of background images are related to the performance of SSD model. This test can lead performance improvement of SSD model or other text-recognizing machine based on deep learning algorithm with high-quality data. SSD model which is re-designed to recognize texts in images and the program developed for creating train data are expected to contribute to improvement of searching system in E-commerce. Suppliers can put less time to register keywords for products and customers can search products with product-details which is written on the catalog.

• Economic and Environmental Geology
• /
• v.45 no.5
• /
• pp.487-502
• /
• 2012

This paper is focused on mineral compositions, microstructures and distributional characters of remained grains in the fault rocks collected from a fault developed in Yongdang-ri, Yangbuk-myeon, Gyeongju City, Korea, using X-ray diffraction (XRD), optical microscope, laser grain size analysis and fractal dimension analysis methods. The exposed fault core zone is about 1.5 meter thick. On the average, the breccia zone is 1.2 meter and the gouge zone is 20cm thick, respectively. XRD results show that the breccia zone consists predominantly of rock-forming minerals including quartz and feldspar, but the gouge zone consists of abundant clay minerals such as chlorite, illite and kaolinite. Mineral vein, pyrite and altered minerals commonly observed in the fault rock support evidence of fault activity associated with hydrothermal alteration. Fractal dimensions based on box counting, image analysis and laser particle analysis suggest that mineral grains in the fault rock underwent fracturing process as well as abrasion that gave rise to diminution of grains during the fault activity. Fractal dimensions(D-values) calculated by three methods gradually increase from the breccia zone to the gouge zone which has commonly high D-values. There are no noticeable changes in D-values in the gouge zone with trend being constant. It means that the bulk-crushing process of mineral grains in the breccia zone was predominant, whereas abrasion of mineral grains in the gouge zone took place by continuous fault activity. It means that the bulk-crushing process of mineral grains in the breccia zone was predominant, whereas abrasion of mineral grains in the gouge zone took place by continuous fault activity. Mineral compositions in the fault zone and peculiar trends in grain distribution indicate that multiple fault activity had a considerable influence on the evolution of fault zones, together with hydrothermal alteration. Meanwhile, fractal dimension values(D) in the fault rock should be used with caution because there is possibility that different values are unexpectedly obtained depending on the measurement methods available even in the same sample.

• Journal of Korean Ophthalmic Optics Society
• /
• v.18 no.2
• /
• pp.179-186
• /
• 2013

Purpose: To compare corneal asphericity, visual acuity (VA), and ocular and corneal higher-order aberrations (HOAs) between myopic refractive surgery and emmetropia groups. Methods: Twenty three subjects ($23.0{\pm}2.5$ years) who underwent myopic refractive surgery and twenty emmetropia ($21.0{\pm}206$ years) were enrolled. The subjects'criteria were best unaided monocular VA of 20/20 or better in both two groups. High and low contrast log MAR visual acuities were measured under photopic and mesopic conditions. Corneal and ocular HOAs were measured using Wavefront Analyzer (KR-1W, Topcon) for 4 mm and 6 mm pupils. Corneal asphericity was taken by topography in KR-1W. Results: There was no significant difference in VA between two groups under either photopic or mesopic conditions. In ocular aberrations, there were significant differences in total HOAs, fourthorder and spherical aberration (SA) for a 6 mm between two groups (p=0.045, p<0.001, and p<0.001, respectively). In corneal aberrations, there was a significant difference in SA for 4 mm (p=0.001) and 6 mm (p<0.001) pupils between two groups and there were statistically significant differences in total HOAs (p<0.001) and fourth-order aberrations (p<0.001) between two groups for a 6 mm pupil. There was a significant correlation in emmetropia between Q-value and SA in ocular aberrations for 4 mm and 6 mm pupils (r=0.442, p=0.004, and r=0.519, p<0.001) and in corneal aberrations for 4 mm and 6 mm pupils (r=0.358, p=0.023, and r=0.646, p<0.001). No significant correlations were found between Q-value and SA in refractive surgery group. Conclusions: VA in myopic refractive surgery is better than or similar to emmetropia. Nevertheless, the more increasing pupil size is, the more increasing aberrations are. Thus, it could have an influence on the quality of vision at night.

• Proceedings of the Korea Association of Crystal Growth Conference
• /
• 1996.06a
• /
• pp.258-292
• /
• 1996

In the last year great interest appears to YBCO thin films preparation on different substrate materials. Preparation of epitaxial film is a very difficult problem. There are many requirements to substrate materials that must be fullfilled. Main problems are lattice mismatch (misfit) and similarity of structure. From paper [1] or follows that difference in interatomic distances and angles of substrate and film is mire important problem than similarity of structure. In this work we present interatomic distances and angle relations between substrate materials belonging to ABCO4 group (where A-Sr or Ca, B-rare earth element, C-Al or Ga) of different orientations and YBCO thin films. There are many materials used as substrates for HTsC thin films. ABCO4 group of compounds is characterized by small dielectric constants (it is necessary for microwave applications of HTsC films), absence of twins and small misfit [2]. There most interesting compounds CaNdAlO4, SrLaAlO4 and SrLaGaO4 were investigated. All these compounds are of pseudo-perovskite structure with space group 14/mmm. This structure is very similar to structure of YBCO. SLG substrate has the lowest misfit (0.3%) and dielectric constant. For preparation of then films of substrates of this group of compound plane of <100> orientation are mainly used. Good quality films of <001> orientations are obtained [3]. In this case not only a-a misfit play role, but c-3b misfit is very important too. Sometimes, for preparation of thin films substrates of <001> and <110> orientations were manufactured [3]. Different misfits for different YBCO faces have been analyzed. It has been found that the mismatching factor for (100) face is very similar to that for (001) face so there is possibility of preparation of thin films on both orientations. SrLaAlO4(SLA) and SrLaGaO4(SLG) crystals of general formula ABCO4 have been grown by the Czochralski method. The quality of SLA and SLG crystals strongly depends on axial gradient of temperature and growth and rotation rates. High quality crystals were obtained at axial gradient of temperature near crystal-melt interface lower than 50℃/cm, growth rate 1-3 mm/h and the rotation rate changing from 10-20pm[4]. Strong anisotropy in morphology of SLA and SLG single crystals grown by the Czochralski method is clearly visible. On the basics of our considerations for ABCO4 type of the tetragonal crystals there can appear {001}, {101}, and {110} faces for ionic type model [5]. Morphology of these crystals depend on ionic-covalent character of bonding and crystal growth parameters. Point defects are observed in crystals and they are reflected in color changes (colorless, yellow, green). Point defects are detected in directions perpendicular to oxide planes and are connected with instability of oxygen position in lattice. To investigate facets formations crystals were doped with Cr3+, Er3+, Pr3+, Ba2+. Chromium greater size ion which is substituted for Al3+ clearly induces faceting. There appear easy {110} faces and SLA crystals crack even then the amount of Cr is below 0.3at.% SLG single crystals are not so sensitive to the content of chromium ions. It was also found that if {110} face appears at the beginning of growth process the crystal changes its color on the plane {110} but it happens only on the shoulder part. The projection of {110} face has a great amount of oxygen positions which can be easy defected. Pure and doped SLA and SLG crystals measured by EPR in the<110> direction show more intensive lines than in other directions which allows to suggest that the amount of oxygen defects on the {110} plane is higher. In order to find the origin of colors and their relation with the crystal stability, a set of SLA and SLG crystals were investigated using optical spectroscopy. The colored samples exhibit an absorption band stretching from the UV absorption edge of the crystal, from about 240 nm to about 550 m. In the case of colorless sample, the absorption spectrum consists of a relatively weak band in the UV region. The spectral position and intensities of absorption bands of SLA are typical for imperfection similar to color centers which may be created in most of oxide crystals by UV and X-radiation. It is pointed out that crystal growth process of polycomponent oxide crystals by Czochralski method depends on the preparation of melt and its stoichiometry, orientation of seed, gradient of temperature at crystal-melt interface, parameters of growth (rotation and pulling rate) and control of red-ox atmosphere during seeding and growth (rotation and pulling rate) and control of red-ox atmosphere during seeding and growth. Growth parameters have an influence on the morphology of crystal-melt interface, type and concentration of defects.

• Horticultural Science & Technology
• /
• v.32 no.4
• /
• pp.448-453
• /
• 2014

Influences of shading and irrigation in summer hot pepper cultivation on the plant growth and mesophyll tissue were investigated. Hot pepper plants were exposed to three shade levels (0, $30{\pm}5$ and $80{\pm}5%$) and irrigated or non-irrigated in greenhouse condition. Plant height and leaf area were highest in 30% shading and stem diameter and fresh and dry weights were highest in no shading. Plant growth was better in rain shelters with irrigation than in those without irrigation. The numbers of hot pepper fruits in the beginning of harvest were 49 in rain shelters without irrigation and shading, 22 in those with irrigation and without shading, 5 in those without irrigation with 30% shading, and 1 in those with irrigation and 30% shading. However, 80% shading showed lowest flower number and flower abscission, resulting in no fruit set, regardless of irritation. This is because carbohydrate translocation from leaves to reproductive organs may be not enough for developing fruits due to the lack of sunlight. The yield of hot pepper tended to be higher in rain shelter with irrigation than in those without irrigation. In optical microscopy observation, the thickness and development of mesophyll tissues decreased as increasing the degree of shading but no effect of irrigation on mesophyll tissues was observed. When stomata were observed with scanning electron microscope (SEM), the shape of stomata was normal but tissues surrounding stomata were slightly wrinkled in plants grown under 30% shading. The large number of abnormal stomata and wrinkled leaves was observed among plants grown in rain shelters with 80% shading. In plants grown in rain shelters without irrigation, tissues surrounding stomata were wrinkled and 10-20% decrease in the number of stomata was observed. Therefore, in hot pepper cultivation in summer with high temperature, shading was not effective for fruit yield and mesophyll tissue development; if shading is unavoidable, high degree of shading is not advisable. Further studies are needed for appropriate cultivar selection and environment-control techniques in hot pepper cultivation in summer with high temperature.

• Restorative Dentistry and Endodontics
• /
• v.35 no.3
• /
• pp.164-172
• /
• 2010

During a composite resin restoration, an anticipating contraction gap is usually tried to seal with low-viscosity resin after successive polishing, etching, rinsing and drying steps, which as a whole is called rebonding procedure. However, the gap might already have been filled with water or debris before applying the sealing resin. We hypothesized that microleakage would decrease if the rebonding agent was applied before the polishing step, i.e., immediately after curing composite resin. On the buccal and lingual surfaces of 35 extracted human molar teeth, class V cavities were prepared with the occlusal margin in enamel and the gingival margin in dentin. They were restored with a hybrid composite resin Z250 (3M ESPE, USA) using an adhesive AdperTM Single Bond 2 (3M ESPE). As rebonding agents, BisCover LV (Bisco, USA), ScotchBond Multi-Purpose adhesive (3M ESPE) and an experimental adhesive were applied on the restoration margins before polishing step or after successive polishing and etching steps. The infiltration depth of 2% methylene blue into the margin was measured using an optical stereomicroscope. The correlation between viscosity of rebonding agents and mciroleakage was also evaluated. There were no statistically significant differences in the microleakage within the rebonding procedures, within the rebonding agents, and within the margins. However, when the restorations were not rebonded, the microleakage at gingival margin was significantly higher than those groups rebonded with 3 agents (p < 0.05). The difference was not observed at the occlusal margin. No significant correlation was found between viscosity of rebonding agents and microleakage, except very weak correlation in case of rebonding after polishing and etching at gingival margin.

• Journal of Korean Ophthalmic Optics Society
• /
• v.13 no.3
• /
• pp.79-87
• /
• 2008

Purpose: This study measured Y-intercept that means a fixation disparity, X-intercept that means a associated phoria and slope of a fixation disparity curve (FDC), which are variables of dissociated phoria and the FDC. We searched for the result value and examined the direction, distribution pattern and a variable that give an most affect on dissociated phoria and the FDC at distance and near. Also this study examined that there were statistically significant differences between distance and near, comparing the result value. Methods: We measured the dissociated phoria and the fixation disparity for 51 subjects at distance and near in June, 2007. All subjects ranged from 20 to 25 years of age (average 21.72${\pm}$1.88 years old) and had no eye disease. At distance the dissociated phoria measured with the distance MIM card (muscle imbalance measure card, Bernell co., USA), and the fixation disparity measured with the modified Mallett Far Unit (Bernell co., USA). At near the dissociated phoria measured with the near MIM card (muscle imbalance measure card, Bernell co., USA), and the fixation disparity measured with the Wesson fixation disparity card (American Optical co., USA). Results: The percentage distribution of types of fixation disparity curves was that at distance prevalence of Type I (74.6%) was the highest, followed by Type IV (17.6%) and Type II (3.9%), Type III (3.9%) and that at near prevalence of Type I (53.0%) was the highest, followed by Type III (29.4%), Type IV (13.7%) and Type II (3.9%). 2. There were significantly correlation in dissociated phoria, fixation disparity (Y-intercept) and associated phoria (X-intercept). 3. The fixation disparity at distance was most affected by associated phoria (X-intercept) (p=0.000). The distance dissociated phoria was most affected by fixation disparity (Y-intercept) (p=0.342), but the influence was weak. 4. The fixation disparity at near was most affected by associated phoria (X-intercept) (p=0.000). The near also dissociated phoria was most affected by associated phoria (X-intercept) (p=0.009). The result that compared the each variables with the same variables at distance and near had statistically significant on paired t-test for among dissociated phoria (t=7.529, p=0.000), X-intercept (t=5.860, p=0.000), the Y-intercept (t=4.640, p=0.000) but slope of the FDC did not differ significant (t=1.336 p=0.188). Conclusions: Relationship of fixation disparity and Heterophoria had close correlation at distance and near.

• Korean Journal of Remote Sensing
• /
• v.37 no.5_1
• /
• pp.989-998
• /
• 2021

In the utilization of optical satellite imagery, which is greatly affected by clouds, periodic composite technique is a useful method to minimize the influence of clouds. Recently, a technique for selecting the optimal pixel that is least affected by the cloud and shadow during a certain period by directly inputting cloud and cloud shadow information during period compositing has been proposed. Accurate extraction of clouds and cloud shadowsis essential in order to derive optimal composite results. Also, in the case of an surface targets where spectral information is important, such as crops, the loss of spectral information should be minimized during cloud-free compositing. In thisstudy, clouds using two spectral indicators (Haze Optimized Tranformation and MeanVis) were used to derive a detection technique with low loss ofspectral information while maintaining high detection accuracy of clouds and cloud shadowsfor cabbage fieldsin the highlands of Gangwon-do. These detection results were compared and analyzed with cloud and cloud shadow information provided by Sentinel-2A/B. As a result of analyzing data from 2019 to 2021, cloud information from Sentinel-2A/B satellites showed detection accuracy with an F1 value of 0.91, but bright artifacts were falsely detected as clouds. On the other hand, the cloud detection result obtained by applying the threshold (=0.05) to the HOT showed relatively low detection accuracy (F1=0.72), but the loss ofspectral information was minimized due to the small number of false positives. In the case of cloud shadows, only minimal shadows were detected in the Sentinel-2A/B additional layer, but when a threshold (= 0.015) was applied to MeanVis, cloud shadowsthat could be distinguished from the topographically generated shadows could be detected. By inputting spectral indicators-based cloud and shadow information,stable monthly cloud-free composited vegetation index results were obtained, and in the future, high-accuracy cloud information of Sentinel-2A/B will be input to periodic cloud-free composite for comparison.

• Korean Journal of Remote Sensing
• /
• v.39 no.6_1
• /
• pp.1211-1224
• /
• 2023

Clouds cause many difficult problems in observing land surface phenomena using optical satellites, such as national land observation, disaster response, and change detection. In addition, the presence of clouds affects not only the image processing stage but also the final data quality, so it is necessary to identify and remove them. Therefore, in this study, we developed a new cloud detection technique that automatically performs a series of processes to search and extract the pixels closest to the spectral pattern of clouds in satellite images, select the optimal threshold, and produce a cloud mask based on the threshold. The cloud detection technique largely consists of three steps. In the first step, the process of converting the Digital Number (DN) unit image into top-of-atmosphere reflectance units was performed. In the second step, preprocessing such as Hue-Value-Saturation (HSV) transformation, triangle thresholding, and maximum likelihood classification was applied using the top of the atmosphere reflectance image, and the threshold for generating the initial cloud mask was determined for each image. In the third post-processing step, the noise included in the initial cloud mask created was removed and the cloud boundaries and interior were improved. As experimental data for cloud detection, CAS500-1 L2G images acquired in the Korean Peninsula from April to November, which show the diversity of spatial and seasonal distribution of clouds, were used. To verify the performance of the proposed method, the results generated by a simple thresholding method were compared. As a result of the experiment, compared to the existing method, the proposed method was able to detect clouds more accurately by considering the radiometric characteristics of each image through the preprocessing process. In addition, the results showed that the influence of bright objects (panel roofs, concrete roads, sand, etc.) other than cloud objects was minimized. The proposed method showed more than 30% improved results(F1-score) compared to the existing method but showed limitations in certain images containing snow.

• Conservation Science in Museum
• /
• v.31
• /
• pp.21-37
• /
• 2024

This paper sought to garner an understanding of Daegaya glass culture by observing the micro-structure and analyzing the chemical composition of 43 glass beads excavated from the No.73 and 74 ancient tombs in Jisan-dong, Goryeong, which are estimated to have a central age of 5th century CE. The visible characteristics and micro-structure of these artifacts were observed with a optical microscope and an scanning electron microscope, while their chemical composition was analyzed with an energy-dispersing spectrometer attached to the scanning electron microscope. As a result, the glass beads of Jisan-dong, Goryeong were identified to have been formed using various methods such as drawing, casting, and folding techniques, with the majority molded by the drawing technique. In terms of chemical composition, 32.6% were in the potash glass group and 67.4% in the soda glass group, with the latter divided into various fluxes such as high alumina glass, netron glass, and plant ash glass. Compared to Baekje's cultural region in the same age, the composition of these ancient glass artifacts demonstrates a high share of the potash glass group. This shows that, despite the shift from the potash glass group to the soda glass group in ancient Korean glass culture, glass composition differs from region to region or depending on the cultural sphere of influence. In the soda glass group, high-alumina glass comprised 23.3%, natron glass 43.0%, and plant ash glass was 1.2%. Among them, the main type of Korean soda glass is high-alumina glass, as natron glass and plant ash glass are known to have appeared later, but the results of scientific analysis of the glass beads excavated in Jisan-dong can be expected to provide important clues about the inflow and transformation of ancient glass on the Korean Peninsula. In the No. 73, 74, and 74-1 ancient tombs, which were found to have been built in chronological order by the excavation survey, the glass beads showed only slight variations depending on their production period. Nonetheless, the chemical composition of glass is deemed to have a close correlation to color.