• KIPS Transactions on Software and Data Engineering
• /
• v.12 no.5
• /
• pp.217-228
• /
• 2023

Video instance segmentation is an intelligent visual task with high complexity because it not only requires object instance segmentation for each image frame constituting a video, but also requires accurate tracking of instances throughout the frame sequence of the video. In special, human instance segmentation in drama videos has an unique characteristic that requires accurate tracking of several main characters interacting in various places and times. Also, it is also characterized by a kind of the class imbalance problem because there is a significant difference between the frequency of main characters and that of supporting or auxiliary characters in drama videos. In this paper, we introduce a new human instance datatset called MHIS, which is built upon drama videos, Miseang, and then propose a novel video data augmentation method, CDVA, in order to overcome the data imbalance problem between character classes. Different from the previous video data augmentation methods, the proposed CDVA generates more realistic augmented videos by deciding the optimal location within the background clip for a target human instance to be inserted with taking rich spatio-temporal context embedded in videos into account. Therefore, the proposed augmentation method, CDVA, can improve the performance of a deep neural network model for video instance segmentation. Conducting both quantitative and qualitative experiments using the MHIS dataset, we prove the usefulness and effectiveness of the proposed video data augmentation method.

• Trans-
• /
• v.12
• /
• pp.51-79
• /
• 2022

This study investigates the methods for deriving colors which can serve as a reference to users such as designers and or contents creators who search for online images from the web portal sites using specific words for color planning and more. Two experiments were conducted in order to accomplish this. Digital scenery photos within the geographic scope of Korea were downloaded from web portal sites, and those photos were studied to find out what colors were used to describe daytime and nighttime. Machine learning was used as the study methodology to classify colors in daytime and nighttime, and KSCA was used to derive the color frequency of daytime and nighttime photos and to compare and analyze the two results. The results of classifying the colors of daytime and nighttime photos using machine learning show that, when classifying the colors by 51~100%, the area of daytime colors was approximately 2.45 times greater than that of nighttime colors. The colors of the daytime class were distributed by brightness with white as its center, while that of the nighttime class was distributed with black as its center. Colors that accounted for over 70% of the daytime class were 647, those over 70% of the nighttime class were 252, and the rest (31-69%) were 101. The number of colors in the middle area was low, while other colors were classified relatively clearly into day and night. The resulting color distributions in the daytime and nighttime classes were able to provide the borderline color values of the two classes that are classified by brightness. As a result of analyzing the frequency of digital photos using KSCA, colors around yellow were expressed in generally bright daytime photos, while colors around blue value were expressed in dark night photos. For frequency of daytime photos, colors on the upper 40% had low chroma, almost being achromatic. Also, colors that are close to white and black showed the highest frequency, indicating a large difference in brightness. Meanwhile, for colors with frequency from top 5 to 10, yellow green was expressed darkly, and navy blue was expressed brightly, partially composing a complex harmony. When examining the color band, various colors, brightness, and chroma including light blue, achromatic colors, and warm colors were shown, failing to compose a generally harmonious arrangement of colors. For the frequency of nighttime photos, colors in approximately the upper 50% are dark colors with a brightness value of 2 (Munsell signal). In comparison, the brightness of middle frequency (50-80%) is relatively higher (brightness values of 3-4), and the brightness difference of various colors was large in the lower 20%. Colors that are not cool colors could be found intermittently in the lower 8% of frequency. When examining the color band, there was a general harmonious arrangement of colors centered on navy blue. As the results of conducting the experiment using two methods in this study, machine learning could classify colors into two or more classes, and could evaluate how close an image was with certain colors to a certain class. This method cannot be used if an image cannot be classified into a certain class. The result of such color distribution would serve as a reference when determining how close a certain color is to one of the two classes when the color is used as a dominant color in the base or background color of a certain design. Also, when dividing the analyzed images into several classes, even colors that have not been used in the analyzed image can be determined to find out how close they are to a certain class according to the color distribution properties of each class. Nevertheless, the results cannot be used to find out whether a specific color was used in the class and by how much it was used. To investigate such an issue, frequency analysis was conducted using KSCA. The color frequency could be measured within the range of images used in the experiment. The resulting values of color distribution and frequency from this study would serve as references for color planning of digital design regarding natural scenery in the geographic scope of Korea. Also, the two experiments are meaningful attempts for searching the methods for deriving colors that can be a useful reference among numerous images for content creator users of the relevant field.

• Journal of the Korean Institute of Traditional Landscape Architecture
• /
• v.28 no.1
• /
• pp.113-119
• /
• 2010

This study examined figurative characters of rice terraces, and is intended to seek figurative possibility of a natural scene by selecting Gyeongnam Area Namhae Gacheon rice terraces, Sancheong Imcheon rice terraces and Hamyang Macheon rice terraces through an analysis of the target land. As a theoretical background, studies of a natural scene and figuration-related literature were looked into, and for figuring a natural scene on the basis of the results, the character of the topography of rice terraces was divided into 9 items as follows by generalizing the inspiration obtained from the natural scenic character and spatial character of rice terraces represented from the topography: (1) Inclination showing the degree of the slope of rice terraces, (2) The bending showed by unevenness of the edge line of rice terraces, (3) The degree of parallelism of the up-and-down edge line, (4) The size of the width of the section of each rice terrace, (5) The area of the total rice terraces, (6) Step gap that indicates the difference of high and low of the up-and-down small rice terraces, (7) Figuration as an image represented by the external form of the total rice terraces, (8) Naturalness and artificiality kept by the total scene, (9) Softness of cornerstones or soil that forms small rice terraces, an atmosphere produced by the total scenes, etc. The results of analysis made through GIS, with rice terraces of Gyeongnam Namhae, Sancheong and Hamyang Area as the target land, are as follows. First, the steeper the inclination is, the higher the step difference of cornerstones becomes, and the steeper the inclination, the narrower the area and width become, so that it could be known that, in rice terraces, inclination is proportional to step difference and the area is inversely proportional to the width. Second, it is considered that line showed in rice terraces, area elements and reiteration, repeated changes, and changes in the scene affected by the surrounding environment including those according to season can be converted into a variety of images according to an individual disposition. Third, it was decided that it is possible to figure repetitive composition and directivity, motility, order and change, etc. into a common type drawn from each target land. This paper, as a study of basic materials for expressing figuration in an environment field, investigated and analyzed the surrounding environment of the target land, and is significant in that it attempted the possibility of being able to generalize the environmental beauty of the target land as one figurative element. More precise studies are expected to be made later.

• Journal of Chest Surgery
• /
• v.42 no.1
• /
• pp.79-86
• /
• 2009

Background: Background: Computed tomography (CT) is the main tool for detecting abnormalities of the thoracic aorta, but conventional CT only shows the cross-sectional images. These CT images have some limitations fo accuratly measuring the thoracic aortic diameters at various levels. Multidetector computed tomography (MDCT) overcomes these limitations. We measured the thoracic aortic diameter perpendicular to the loop-shaped thoracic aortic course and this was studied in relation to age, gender, height, weight, the body surface area, the body mass index and the presence of hypertension. Material and Method: Thirty hundred thirty one patients (males: 141 patients and females: 190 patients) who had no abnormalities of the thoracic aorta were investigated using MDCT aortography. They were divided into three age categories: 20~39 years old, 40~59 years old and over age 60. The image was reformed with multiplanar reconstruction and the diameter of the aorta was measured perpendicular to the aortic course at 5 anatomic segments. Level A was the mid-ascending aorta, level B was the distal ascending aorta, level C was the aortic arch, level D was the aortic isthmus and level E was the mid-descending aorta. Result: The mean age was 49.5 years old for males and 54.9 years old for females (p<0.05). The mean diameter of the thoracic aorta at level A was 31.1 mm, that at level B was 30.2 mm, that at level C was 26.5 mm, that at level D was 24.0 mm and that at level E was 22.6 mm. The diameters at all the levels were gradually increased with age. Hypertensive patients had larger diameters than did the non-hypertensive population. There was a positive correlation between the ascending aortic diameter (levels A&B) and height and the body surface area, but there were no statistical differences at the aortic arch (level C) and the descending aorta (levels D&E). There were no statistical differences of the weight and body mass index at all levels. Conclusion: The diameters of the thoracic aortas were directly correlated with gender, age and hypertension. Height and the body surface area were only correlated with the ascending aorta. Weight and the body mass index have no statistical difference at all levels. We measured the age related thoracic aortic diameters and the upper normal limits and we provide this data as reference values for the thoracic aortic diameter in the Korean population.

• Tuberculosis and Respiratory Diseases
• /
• v.65 no.4
• /
• pp.269-276
• /
• 2008

Background: Although the prevalence of tuberculosis infections (PTBI) is one of the basic epidemiologic indices, no survey has been carried out since 1995 because the nation-wide tuberculosis prevalence survey was changed to a surveillance system. Subjects without a BCG scar are examined in a tuberculin survey. However, it is very difficult to select these subjects under high vaccination coverage. It is important to evaluate the impact of BCG vaccinations on the tuberculin response and estimate the PTBI regardless of the BCG vaccination status. Methods: A nation-wide, school-based cross-sectional tuberculin survey was carried out among first graders in elementary school in 2006. A total of 5,148 children in 40 schools were selected by quota sampling. Tuberculin testing with 0.1 ml of two tuberculin units of PPD RT23 was carried out on 4,018 children. The maximum transverse diameter of induration was measured 48 to 72 hours later. The presence of a BCG scar was checked separately. Results: There were no BCG scars in 6.3% of the subjects. The mean induration size of tuberculin testing was $3.7{\pm}4.4mm$, which included 1,882 (46.8%) subjects with an induration size of 0 mm. The PTBI was 10.9% (439 subjects) using a cut-off point of ${\geq}10mm$ (conventional method). The annual risk of tuberculosis infections (ARTI) was 1.9% when the mean age of the subjects was assumed to be 6 years. There was no difference in the PTBI according to the presence or absence of a BCG scar [11.2% vs 7.6% (OR: 1.54, 95% CI: 0.98~2.43)]. Using a mirror image technique with 16 mm as the cut-off point, the PTBI and ARTI had decreased to 2.4% and 0.4% respectively. Conclusion: PTBI and ARTI, as estimated by conventional methods, appear to be high among BCG vaccinated children. A mirror image technique is more suitable for estimating the indices in a country with an intermediate burden of tuberculosis than the conventional method.

• Tuberculosis and Respiratory Diseases
• /
• v.48 no.2
• /
• pp.191-202
• /
• 2000

Background: Bronchial asthma is characterized by airway hyperresponsiveness(BHR) and inflammation. The cyclooxygenase(COX) is believed to be one of the important enzymes in these inflammatory reactions. Recently, the COX was divided into two isoforms, COX1 and COX2. COX2 is induced by lipopolysaccharide and some cytokines at the inflammation site. Prostaglandin E2(PGE2), produced from COX2, may affect airway inflammation. The purpose of this study is to evaluate the effect of COX2 inhibitor on COX2 expression, plasma PGE2, airway resistance and histologic finding in an animal asthma model. Methods : Sprague-Dawley rats were divided into 3 groups. The normal control group did not receive any treatment, but the asthma control group was sensitized by ovalbumin but not treated with the COX2 inhibitor(nimesulide, Mesulid$^{(R)}$). The treatment group was sensitized and treated with nimesulide. Specific airway resistance(sRaw) before and after nimesulide ingestion was investigated. The PGE2 level in the plasma was examined and COX2 immunogold-silver stain on lung tissue was performed. Results: sRaw and eosionophilic infiltration on airway, which increased in the asthma control group, was compared to normal control(p=0.014). However, there was no difference in eosinophilic infiltration between asthma control and treatment groups(p=0.408) and no difference in COX2 expression on bronchiolar epithelium among the three groups. Plasma PGE2 levels were not statically different among the three groups. Conclusion: The role of COX2 in the allergen-induced BHR was not significant The effect of nimesulide was not observed on BHR, COX2 expression, and plasma PGE2 level. Therefore, COX2 may not be a major substance of allergic asthma.

• Journal of Chest Surgery
• /
• v.34 no.7
• /
• pp.534-538
• /
• 2001

Background: Knowledge of size and morphology of the normal trachea is important for airway management and tracheal reconstruction. Conventional radiography is a simple method used to measure the tracheal diameter, but it is not accurate because of the artifacts related to image magnification and overlapping by the shoulder. The purpose of this study was to provide the normal values of the tracheal size and anatomy in Korean adults using Computerized Topography. Material and Method: There were 43 men and 34 women included in this study. They were divided into three age groups(group 1, 20-39 years ; group 2, 40-59 yeas , groups 3, $\geq$60 years). The anteroposterior and transverse diameters and cross - sectional areas of the trachea were measured at the level of the thoracic inlet(Level 1) and the aortic arch(Level 2). These values obtained at each level were compared between age groups and sexes. Result: In 43 men, the anteroposterior / transverse diameters(mean SD in millimeters) of the trachea at levels 1 and 2 were 19.95$\pm$2.99 / 17.72$\pm$2.13 and 19.77$\pm$2.57 / 18.02$\pm$2.19, respectively. In 34 women, those values at levels 1 and 2 were 15.56$\pm$2.12 / 14.18$\pm$2.07 and 15.35$\pm$1.82 / 15.00$\pm$1.60, respectively. At both levels, the anteroposterior and transverse diameters were significantly greater in men than in women (p<0.05). The cross-sectional area of the trachea at levels 1 and 2 were 279.14$\pm$61.37 / 281.93$\pm$63.97 $\textrm{mm}^2$ in men and 173.29$\pm$35.81 / 181.88$\pm$34.74 in women, respectively. They also showed significantly greater values in men than in women(P<0.05). There was no significant difference in diameters and cross-sectional areas of the trachea between age groups. Conclusion: There are significant differences in the internal diameter and cross- sectional area of the trachea between men and women in normal Korean adults, while the age difference was insignificant. We believed CT is a relatively accurate and safe way to measure the internal diameter and cross-sectional areas of the trachea.

• The Korean Journal of Nuclear Medicine Technology
• /
• v.16 no.1
• /
• pp.27-33
• /
• 2012

Purpose: Most of diagnosis in the pediatric hydronephrosis patients have been performed $^{99m}Tc$-DMSA renal scan. Then the region of interest (ROI) is set for comparative analysis of uptake ratio in left-right kidney after acquiring the image. But if the equipment set an automatic ROI, the ROI could include expanded renal pelvis due to hydronephrosis and the uptake ratio of left-right kidney will be incorrect result. Therefore this study compared both ROIs including expanded renal pelvis and excluding renal pelvis through experiment using normal kidney phantom and expanded renal pelvis phantom and suggested setting method of improved ROI. In addition, this study have been helped by readout doctor for investigate distinction radiopharmaceutical uptake between renal cortex and remained urine by expanded renal pelvis. Materials and Methods: The both of renal phantoms were filled with water and shacked with $^{99m}TcO_4$ 111 MBq. In order to describe the expanded renal pelvis, the five latex balloon were all filled with 10 mL water and each of balloon was mixed with $^{99m}TcO_4$ 18.5, 37, 55.5, 74, 92.5 MBq. And we made phantom with fixed $^{99m}TcO_4$activity of 37 MBq and mixed water 5, 10, 15, 20, 25 mL in each balloon. The left kidney was fixed its shape and the right kidney was modified like as hydronephrosis kidney by attached the latex balloons. And the acquiring counts were 2 million. After acquisition, we compared the image of ROI with Expanded renal pelvis and the image of ROI without renal pelvis for analyzing difference in the uptake ratio of left-right kidney and for reproducibility, set the ROI 5 times in the same images. Patients were injected $^{99m}Tc$-DMSA 1.5~1.9 MBq/kg and scanned 3 to 4 hours after injection. The each of 3 skillful radio technologists performed the comparing estimation by setting ROI. To determine statistical significance between two data, SPSS (ver. 17) Wilcoxon Signed Ranks Test was used. Results: As a result of renal phantom's experiment, we compared with average of counts Background (BKG) ratios in the setting of ROI including expanded renal pelvis and setting of excluding expanded renal pelvis. Therefore, they can obtain changed counts and changed ratios. Patient also can obtain same results. In addition, the radiopharmaceutical uptake in expanded renal pelvis was come out the remained urine that couldn't descend to ureter by the help of readout doctor. Conclusion: As above results, the case of setting ROI including expanded renal pelvis was more abnormally increasing uptake ratio than the case of setting ROI excluding expanded renal pelvis in analysis the uptake ratio in left-right kidney of hydronephrosis. Because of the work convenience and prompted analysis, the automatic ROI is generally used. But in case of the hydronephrosis study, we should set the manual ROI without expanded renal pelvis for an accurate observation of the uptake ratio of left-right kidney since the radiopharmaceutical uptake in expanded renal pelvis is the remained urine.

• 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.

• Journal of Chest Surgery
• /
• v.43 no.6
• /
• pp.596-601
• /
• 2010

Background: This study was designed to evaluate the efficacy of a mixture of hyaluronic acid and sodium carboxymethyl cellulose ($Guardix-sol^{(R)}$) on experimental pericardial adhesion. Material and Method: Thirty rats were divided into 2 groups of 15 rats each and pericardial mesothelial injury was induced during surgery by abrasion. In the control group, blood and normal saline were administered into pericardium; in the test group, blood and HA-CMC solution were administered. Pericardial adhesions were evaluated at 2 weeks (n=5), 4 weeks (n=5), and 6 weeks (n=5) after surgery. The severity of adhesions was graded by macroscopic examination, and the adhesion tissue thickness was analyzed microscopically with Masson trichrome stain and an image processing program. Result: The test group had significantly lower macroscopic adhesion scores ($2.9{\pm}0.6$ : $3.9{\pm}0.4$, p<0.000) compared with the control group. For microscopic adhesion tissue thickness, the test group had lower scores compared with the control group, but this difference was not statistically significant ($91.73{\pm}49.91$ : $117.67{\pm}46.4$, p=0.106). Conclusion: We conclude that an HA-CMC solution ($Guardix-sol^{(R)}$) reduces the formation of pericardial adhesions in this animal model.