• Journal of Korea Multimedia Society
• /
• v.23 no.4
• /
• pp.539-548
• /
• 2020

Deep learning technology is currently being used and applied in many different fields. Convolution neural network (CNN) is a method of artificial neural networks in deep learning, which is commonly used for analyzing different types of images through classification. In the conventional classification of histopathology images of prostate carcinomas, the rating of cancer is classified by human subjective observation. However, this approach has produced to some misdiagnosing of cancer grading. To solve this problem, CNN based classification method is proposed in this paper, to train the histological images and classify the prostate cancer grading into two classes of the benign and malignant. The CNN architecture used in this paper is based on the VGG models, which is specialized for image classification. However, color normalization was performed based on the contrast enhancement technique, and the normalized images were used for CNN training, to compare the classification results of both original and normalized images. In all cases, accuracy was over 90%, accuracy of the original was 96%, accuracy of other cases was higher, and loss was the lowest with 9%.

• Journal of Korea Multimedia Society
• /
• v.11 no.9
• /
• pp.1245-1257
• /
• 2008

The establishment of a production pipeline imported 30 computer graphics is suggested in this paper to improve the efficiency of existing production pipeline of clay animation. The point is that the process of building clay characters that remains labor intensive among the existing procedures is replaced by the process of creating computer generated characters. In order to create characters out of clay by means of 30 computer graphics, a diffuse map and displacement map are made of an oil-based clay according to the UVW coordination of polygon modeling, which is the same color and kind of clay used to make a clay character. In addition, a panoramic HDRI recording system is developed to record the lighting information of shooting environment for miniature sets, which is imported in 3D computer graphic tools as digital light source. On account of the new production pipeline, a hyper realistic rendering image can be produced, and at the same time it improves the traditional pipeline of stop motion animation that is know-how based procedure of a complete artist by the engineering approach to the automatic process.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.41 no.5
• /
• pp.602-609
• /
• 2016

In a smart society, politicians analyze the big data of voters to build a favorable political positions. In other words, a variety of digital footprints uploaded in SNS or Internet are used to set the election strategies and political directions. In comparison, it is difficult for voters to extract intention information about how politicians are performing a political acts. Therefore, it is important that voters need to analyze what intention of politicians are like for two way interaction between voters and politicians. For this, in this paper, we want to do the identification by analyzing IT technologies to narrative styles of politicians who pursue relative advantages or gains compared to other competitors. The experiments will be carried out to identify about what relative advantages compared to other competitors by narrative styles of next presidential candidates who are expected to run into the next presidential election by analyzing the usual audio interviews.

• Journal of Fashion Business
• /
• v.17 no.6
• /
• pp.28-43
• /
• 2013

Recent trends toward the collaborations among various sectors of academia and research areas have brought interests and significances in new activities especially in the fashion and textile areas. One of the collaboration examples is the recent research projects on 3D virtual clothing systems based on the 3D CAD software. The 3D virtual clothing systems provide simulated apparels with high degrees of fidelity in terms of color, texture, and structural details. However, since real fabrics exhibit strong nonlinearity, anisotropy, viscoelasticity, and hysteresis, the 3D virtual clothing systems need fine tuning parameters for the simulation process. In this study, characteristics of silk fabrics, which are woven by using degummed silk and raw silk yarns, are being analyzed and compared. Anisotropic properties may be measured as warp and filling direction properties separately in woven fabrics, such as warp tensile stress or filling bending rigidity. Hysteretic properties may be measured as bending hysteresis or shear hysteresis by using KES measurements. These data provide deformation-force relationships of the fabric specimen. Three-dimensional effects obtained when using these characteristic fabrics are also analyzed. The methods to control the three-dimensional appearance of the sewn fabric specimens when utilizing a programmable microprocessor-based motor device, as prepared in this study, are presented. Based on the physical and mechanical properties measured when using the KES equipment, the property parameters are being into a 3-dimensional virtual digital clothing system, in order to generate a virtual clothing product based on the measured silk fabric properties.

• Journal of the Korea Society of Computer and Information
• /
• v.16 no.8
• /
• pp.49-56
• /
• 2011

Recently, animation contents has become extensively available along with the development of cultural industry. In this paper, we propose a method to analyze a face of animation character and extract 3D rotational information of the face. The suggested method first generates a dominant color model of a face by learning the face image of animation character. Our system then detects the face and its components with the model, and establishes two coordinate systems: base coordinate system and target coordinate system. Our system estimates three dimensional rotational information of the animation character face using the geometric relationship of the two coordinate systems. Finally, in order to visually represent the extracted 3D information, a 3D face model in which the rotation information is reflected is displayed. In experiments, we show that our method can extract 3D rotation information of a character face reasonably.

• The Research Journal of the Costume Culture
• /
• v.25 no.3
• /
• pp.285-300
• /
• 2017

This study analyzes the images of models in womenswear advertisements targeting women in their fifties. The goals of this study are: 1) to investigate beauty ideals for middle-aged women by analyzing models' look age, chronological age, wrinkles, gray hair, hair length, body type, and race; and 2) to explore how ageing is dealt with in advertisements by analyzing the range of bodies shown in advertisements, the color mode of photographs, and the clarity of models' figures in relation to models' look ages. A total of 155 printed advertisements from January 2012 to January 2017 from the brands Daks Ladies, Lebeige, Luciano Choi, PAT, and Zishen were selected for analysis. Womenswear brands targeting middle-aged women reinforce cultural ideals of female beauty that emphasize youth and slenderness. They do this by using thin and slender models, who most often appear to be in their twenties and thirties, and have hair longer than their shoulders. Brands with higher price ranges show a preference for Caucasian models, which reveals that a Caucasian identity is associated with sophistication. In addition, the bodies of models who appear to be in their forties and fifties were concealed by framing photographs mostly above the knees. Older models' features were also obscured via the use of black and white photography, strong lighting and contrast, and digital editing that blurred the boundaries between figures and their backgrounds. These decisions for how to represent models could result in negative self-esteem and a denial of the symptoms of ageing among middle-aged women.

• Journal of the Institute of Convergence Signal Processing
• /
• v.5 no.4
• /
• pp.294-300
• /
• 2004

According to the development of digital media technologies various algorithms for video clip matching have been proposed to match the video sequences efficiently. A large number of video search methods have focused on frame-wise query, whereas a relatively few algorithms have been presented for video clip matching or video shot matching. In this paper, we propose an efficient algorithm to index the video sequences and to retrieve the sequences for video clip query. To improve the accuracy and performance of video sequence matching, we employ the Cauchy function as a similarity measure between histograms of consecutive frames, which yields a high performance compared with conventional measures. The key frames extracted from segmented video shots can be used not only for video shot clustering but also for video sequence matching or browsing, where the key frame is defined by the frame that is significantly different from the previous frames. Experimental results with color video sequences show that the proposed method yields the high matching performance and accuracy with a low computational load compared with conventional algorithms.

• Journal of the Institute of Electronics Engineers of Korea SP
• /
• v.43 no.4 s.310
• /
• pp.66-73
• /
• 2006

Safety is the most important for all traffic management and control technology. This paper focuses on developing a flexible, reliable and real-time processing algorithm which is able to generate signal for the entering vehicle at the joining road through a camera and image processing technique. The images obtained from the camera located beside and upon the road can be used for traffic surveillance, the vehicle's travel speed measurement, predicted arriving time in joining area between main road and joining road. And the proposed algorithm displays the confluence safety signal with red, blue and yellow color sign. The three methods are used to detect the vehicle which is driving in setted detecting area. The first method is the gray scale normalized correlation algorithm, and the second is the edge magnitude ratio changing algorithm, and the third is the average intensity changing algorithm The real-time prototype confluence safety signal generation algorithm is implemented on stored digital image sequences of real traffic state and a program with good experimental results.

• The korean journal of orthodontics
• /
• v.45 no.3
• /
• pp.105-112
• /
• 2015

Objective: A recently developed facial scanning method uses three-dimensional (3D) surface imaging with a light-emitting diode. Such scanning enables surface data to be captured in high-resolution color and at relatively fast speeds. The purpose of this study was to evaluate the accuracy and precision of 3D images obtained using the Morpheus 3D$^{(R)}$ scanner (Morpheus Co., Seoul, Korea). Methods: The sample comprised 30 subjects aged 24.34 years (mean $29.0{\pm}2.5$ years). To test the correlation between direct and 3D image measurements, 21 landmarks were labeled on the face of each subject. Sixteen direct measurements were obtained twice using digital calipers; the same measurements were then made on two sets of 3D facial images. The mean values of measurements obtained from both methods were compared. To investigate the precision, a comparison was made between two sets of measurements taken with each method. Results: When comparing the variables from both methods, five of the 16 possible anthropometric variables were found to be significantly different. However, in 12 of the 16 cases, the mean difference was under 1 mm. The average value of the differences for all variables was 0.75 mm. Precision was high in both methods, with error magnitudes under 0.5 mm. Conclusions: 3D scanning images have high levels of precision and fairly good congruence with traditional anthropometry methods, with mean differences of less than 1 mm. 3D surface imaging using the Morpheus 3D$^{(R)}$ scanner is therefore a clinically acceptable method of recording facial integumental data.

• 한국지구물리탐사학회:학술대회논문집
• /
• 2001.09a
• /
• pp.104-121
• /
• 2001

Recently, the pollution of soil and groundwater becomes a serious social problem, and geophysical exploration methods have been introduced as a remedial investigation method of subsurface. Digital technologies such as personal computer have revolutionized our ability to acquire large volume of data in a short term, and to produce more reliable results for subsurface image. Also, color graphics easily visualizes the survey results in a more understandable manner, and it is widely used for not only characterizing the contaminated subsurface but also monitoring contaminant and remedial process. In this paper, electrical resistivity and electromagnetic (EM) surveys were carried out in order to understand characteristics of waste landfills, and the applicability of geophysical prospecting to site assessment of waste landfill was also tested. According to the result, electrical resistivity and electromagnetic (EM) surveys were effective in estimating distribution of the leachate plume.