• Journal of the Korean Society of Costume
• /
• v.63 no.2
• /
• pp.14-28
• /
• 2013

This study attempts to systematically analyze a red carpet style. I have researched the Academy Awards called representative film awards which symbolizes international fame and the Grammy Awards which is the most prestigious award in the music industry by subdividing into formative elements such as silhouette, color, fabric, pattern, detail, accessory, fashion image, and so on from 2000 to 2012. Firstly, when it comes to silhouette, mermaid silhouette accounts for the highest proportion in the Academy Awards. Compared to this, fit silhouette is shown almost the same percentage as the mermaid silhouette in the Grammy Awards. Secondly, with regard to color, black color has not only the highest percentage but also examples of different unit forms such as various color, showy gradation and single colors. Various colors in the Grammy Awards have similar percentage in comparison with the Academy Awards. Thirdly, in terms of fabric, silky material is often used most, which looks like putting more weight on dresses for the formative elements of clothes. Fourthly, in pattern, patternless dresses are represented by high percentage at both the Academy Awards and Grammy Awards. Dresses with patterns have mild, stylistic elements and geometric designs. The Grammy Awards shows many different unique patterns, color and size, compared to the Academy Awards. Fifthly, in detail, frill and ruffle ornaments are shown most at the Academy Awards and Grammy Awards. Especially in the Grammy Awards, beads ornaments are used most. Sixthly, in accessory, there are many accessories of graceful, elegance styles in the Academy Awards. On the contrary to this, there are many accessories to effect on many performances of large, fancy, unique styles. Seventhly, elegance images of a goddess style among fashion images emerge as fashion of the Academy Awards. In spite of romantic styles in the Grammy Awards, many various images are the same rate as there, which means different appearance of experiment and sensational styles.

• Journal of the Institute of Electronics Engineers of Korea SP
• /
• v.48 no.4
• /
• pp.59-66
• /
• 2011

In this paper, we propose a face detection method based on color constancy and geometrical analysis. With the problem about the various colors of skin under scene illuminant, a color constancy method is applied to input images and geometrical analysis is used to detect face regions. At first, the candidates of face or hair are extracted from the image that a color constancy method is applied to, and are classified by some geometrical criterions. And then, face candidates which have some intersectional regions whose total is over a certain size, with hair candidates are selected as faces. Caltech Face DB was used to compare the performance of our method. Also, performance about scene illuminant was evaluated by images which have some illumination effects. The experiment results show that the proposed face detection method was applicable to various facial images because of high true-positive and low false-negative ration.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.16 no.3
• /
• pp.407-415
• /
• 2012

In this paper, we propose a new stereo matching algorithm by using color information especially for stereo images containing human beings in the applications such as tele-presence system. In the proposed algorithm, we first remove the background regions by using a threshold value for stereo images obtained by stereo camera and then find an initial disparity map and segment a given image into R, G, B and white color components. We also obtain edges in the segmented image and estimate the disparity from the extract boundary regions. Finally, we generate the final disparity map by properly combining the disparity map of each color component. Experiment results show better performance compared with the window based method and the dynamic programing method especially for stereo images with human being.

• KIPS Transactions on Software and Data Engineering
• /
• v.4 no.11
• /
• pp.521-528
• /
• 2015

Road-view object classification methods are mostly influenced by weather and illumination conditions, thus the most of the research activities are based on dataset in clean weathers. In this paper, we present a road-view object classification method based on color segmentation that works for all kinds of weathers. The proposed method first classifies the weather and illumination conditions and then applies the weather-specified color models to find the road traffic signs. Using 5 different features of the road-view images, we classify the weather and light conditions as sunny, cloudy, rainy, night, and backlight. Based on the classified weather and illuminations, our model selects the weather-specific color ranges to generate Gaussian Mixture Model for each colors, Green, Yellow, and Blue. The proposed method successfully detects the traffic signs regardless of the weather and illumination conditions.

• Journal of Fashion Business
• /
• v.25 no.5
• /
• pp.1-24
• /
• 2021

This study aimed to explore the fashion design and style features, and differences in BLACKPINK, TWICE, and Red Velvet girl groups. A total of 469 fashion photos (132 BLACKPINK, 217 TWICE, 120 Red Velvet) focusing on 23 title songs (7 BLACKPINK, 9 TWICE, 7 Red Velvet) were collected. Photo classification work was carried out in accordance with the design analysis criteria and the results were derived by combining statistical analysis and content analysis. BLACKPINK's fashion design characteristics showed a lot of complex colors, shade tone, pure tone, contrast color coordination, stylistic pattern, slit, patchwork, checklist method, sexy and avant-garde images. TWICE's fashion design characteristics included warm color, complex color, tint tone, monotone, contrast color coordination, tone-on-tone, geometric & stylistic patterns, cotton, silk, a combination of the same material, frill, beads, ribbon decoration, blouse, skirt, and many romantic and ethnic images. Red Velvet's fashion design characteristics were a cold color, moderate tone, monotone, cotton, velvet, geometric pattern, zipper, sequins, T-shirt, pants, tie, belt, and many retro and active images. The fashion styles of BLACKPINK, TWICE, and Red Velvet were as follows. BLACKPINK was divided into sexy avant-garde, sexy active, sexy romantic styles. TWICE was divided into romantic active, romantic classical, and romantic ethnic styles. Red Velvet was divided into retro active, retro sexy, and retro avant-garde styles.

• Journal of Broadcast Engineering
• /
• v.13 no.5
• /
• pp.651-661
• /
• 2008

This paper proposes a novel contrast enhancement (CE) algorithm for color images using the modified error diffusion (ED). After conventional color histogram equalization (HE), artifacts such as false contours are produced in the contrast enhanced image. The proposed CE algorithm using the modified ED consists of two parts: CE and ED. In the first part, a low-contrast input image is enhanced by the conventional HE method. In the second part, we use the modified ED algorithm. The inputs of the second part are the average and scaled difference images of the original color input image and the HE image, in which the scaled color difference image is diffused by the ED algorithm. In the proposed algorithm, the modified ED algorithm reduces the artifacts produced in the HE image, and increases the number of color levels. Computer simulations with a number of low-contrast color images show the effectiveness of the proposed CE method in terms of the visual quality as well as the probability mass function. It can be used as a post-processing for CE with simultaneous artifact reduction in various display devices.

• The KIPS Transactions:PartB
• /
• v.17B no.3
• /
• pp.197-206
• /
• 2010

High-quality and low-price digital printing devices are nowadays abused to print or forge official documents and bills. Identifying color laser printers will be a step for media forensics. This paper presents a new method to identify color laser printers with printed color images. Since different printer companies use different manufactural systems, printed documents from different printers have little difference in visual. Analyzing this artifact, we can identify the color laser printers. First, high-frequency components of images are extracted from original images with discrete wavelet transform. After calculating the gray-level co-occurrence matrix of the components, we extract some statistical features. Then, these features are applied to train and classify the support vector machine for identifying the color laser printer. In the experiment, total 2,597 images of 7 printers (HP, Canon, Xerox DCC400, Xerox DCC450, Xerox DCC5560, Xerox DCC6540, Konica), are tested to classify the color laser printer. The results prove that the presented identification method performs well with 96.9% accuracy.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.34 no.12C
• /
• pp.1138-1146
• /
• 2009

Digital cameras adopting a single CCD detector collect image color by subsampling in three color planes and successively interpolating the information to reconstruct full-resolution color images. Therefore, to recovery of a full-resolution color image from a color filter array (CFA) like the Bayer pattern is generally considered as an interpolation issue for the unknown color components. In this paper, we first calculate luminance component value by combining R, G, B channel component information which is quite different from the conventional demosaicing algorithm. Because conventional system calculates G channel component followed by computing R and B channel components. Integrating the obtained gradient edge information and the improved weighting function in luminance component, a new edge sensitive demosaicing technique is presented. Based on 24 well known testing images, simulation results proved that our presented high-quality demosaicing technique shows the best image quality performance when compared with several recently presented techniques.

• Journal of Fashion Business
• /
• v.12 no.4
• /
• pp.114-130
• /
• 2008

The purpose of this study is to investigate the effect of Dot Pattern Size(0.8, 1.8, 2.5, 5, 8), color combination(W/Bk, Bk/Gr, Gr/W), Area-Ratio(Background/Dot, Dot/Background) on wearing dot-printed dresses image. Sets of stimulus and response scales(7 point semantic) were used as experimental materials. The stimuli were 30 color pictures manipulated with the combination of Dot Pattern Size, color combination, and Area-Ratio using computer simulation. The subjects were 180 female undergraduates living in Gyeongnam-do. The data was analyzed by using SPSS program. Analyzing methods were ANOVA and LSD test. Image factor of the stimulus was composed of 5 different components, visibility, chastity, attractiveness, cuteness and feminity. Among them, the visibility and chastity were important. Each dimensional image was affected by dot pattern size, color combination and Area-Ratio. In the visibility image, color combination(W/Bk is the most effective) is more influential, the larger size is effective pattern. In the cuteness and feminity image, area ratio(low-brightness dot pattern is the more effective) is more effective than color combination or dot pattern size. Even the same dot pattern size and area was recognized as different image depending on the area ratio. According to the variation of dot pattern size, color combination and area-ratio, it was investigated that the images for a dress wearer were expressed diversely, were shown differently in image dimensions, and could be produced to different images.

• The Journal of the Korea Contents Association
• /
• v.7 no.2
• /
• pp.109-116
• /
• 2007

To reduce the effect of color distortions on reading colors, it is more desirable to statistically process as many pixels in the individual color region as possible. This process may require segmentation, which usually requires edge detection. However, edges in color codes can be disconnected due to various distortions such as dark current, color cross, zipper effect, shade and reflection, to name a few. Edge linking is also a difficult process. In this paper, k-means clustering was performed on the images where edge detectors failed segmentation. Experiments were conducted on 311 images taken in different environments with different cameras. The primary and secondary colors were randomly selected for each color code region. While segmentation rate by edge detectors was 89.4%, the proposed method increased it to 99.4%. Color recognition was performed based on hue, a*, and b* components, with the accuracy of 100% for the successfully segmented cases.