• 한국컴퓨터정보학회논문지
• /
• 제24권1호
• /
• pp.85-92
• /
• 2019

In this paper, we propose a efficient feature point extraction method that can solve the problem of performance degradation by introducing a preprocessing process when extracting feature points by utilizing the characteristics of 360-degree realistic media. 360-degree realistic media is composed of images produced by two or more cameras and this image combining process is accomplished by extracting feature points at the edges of each image and combining them into one image if they cover the same area. In this production process, however, the stitching process where images are combined into one piece can lead to the distortion of non-seamlessness. Since the realistic media of 4K-class image has higher resolution than that of a general image, the feature point extraction and matching process takes much more time than general media cases.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제16권3호
• /
• pp.894-907
• /
• 2022

The features extracted by convolutional neural networks are more descriptive of images than traditional features, and their convolutional layers are more suitable for retrieving images than are fully connected layers. The convolutional layer features will consume considerable time and memory if used directly to match an image. Therefore, this paper proposes a feature weighting and region integration method for convolutional layer features to form global feature vectors and subsequently use them for image matching. First, the 3D feature of the last convolutional layer is extracted, and the convolutional feature is subsequently weighted again to highlight the edge information and position information of the image. Next, we integrate several regional eigenvectors that are processed by sliding windows into a global eigenvector. Finally, the initial ranking of the retrieval is obtained by measuring the similarity of the query image and the test image using the cosine distance, and the final mean Average Precision (mAP) is obtained by using the extended query method for rearrangement. We conduct experiments using the Oxford5k and Paris6k datasets and their extended datasets, Paris106k and Oxford105k. These experimental results indicate that the global feature extracted by the new method can better describe an image.

• 한국실내디자인학회논문집
• /
• 제14권5호
• /
• pp.106-113
• /
• 2005

New images percolate through human consciousness by the media such as movies, TV programs, and brilliant advertisements. These images reproduce new ' things ' throughout the ' semantic processes ' by those who experience and recognize them. Alvin Toffler describes it as the ' information bomb ' and ' image fragments ' in his talk about the new paradigm of information-oriented era. The increasing number of images and their accelerating rate of appearance imply that images become more momentary, and are evidence that they are transforming entire human life and consciousness. Such awareness means a lot to a designer. Especially, the subject that how modern space-dominating images are related to the structure and materials constituting the space and communicate with human mind will be an important factor in establishing the human-space relationship in the future. Furthermore, the present age overspread with various medium is not the only one privileged of the images that exist within space. They are the results of continuous expansion of existing images, and also process of evolution of space powered by the fusion of images and digital media. Imagified space is a boundary layer of Cyberspace, and the space itself becomes an interface by human recognition and participation. Now, the functional classification of spaces such as ' office, ' ' cafe, ' and ' school ' is meaningless. Whatever it may be, the function of a space is defined by the information it interfaces, and therefore it becomes an interface to information through a large number of images. Based on this idea, we will observe the imagificaiton of space in the form of discussion, and from that, try to understand the phenomenon through the real world examples.

• 한국의상디자인학회지
• /
• 제20권2호
• /
• pp.47-62
• /
• 2018

This study examined recycled fashion design utilizing natural images for application to recent trends. This study is significant in that it presents the possibility of the development of recycled women's wear design by reflecting the characteristics and expressions of natural images and using denim material, which is the most common clothing material for everyday life. The results of this study are summarized as follows. First, five types of women's clothing was produced, pursuing a natural design with a soft and feminine silhouette. Second, the colors used were blue and white, which could represent nature. Third, as for materials, this study used clothing that was to be thrown away: four pairs of denim pants and two denim dresses. This study chose denim cloth, with its great value for reuse, because of the characteristics of the strong and durable fiber, because everyone has more than one article of clothing made from denim and because it can be easily sourced. Fourth, for textile motifs, this study expressed a peaceful natural scenery with tropical animals and plants. In addition, this study further emphasized natural images using transfer media printing. This study has significance in that it presented the possibility of recycled fashion design and expanded the range of utilization using transfer media printing, a dyeing treatment to reduce the environmental burden.

• 한국콘텐츠학회논문지
• /
• 제11권5호
• /
• pp.185-193
• /
• 2011

기업은 자사의 브랜드 이미지를 적절하게 구축하고 소비자에게 효과적으로 어필하기 위해 광고를 비롯한 각종 커뮤니케이션을 통한 홍보 수단을 동원한다. 방송 매체별 이용자 비율을 살펴보면, 여전히 텔레비전이 라디오, 신문, 잡지, 인터넷, DMB 등의 다른 주요 매체들에 비해 접촉률과 관심도면에서 월등함을 보여주고 있다. 이렇듯이 소비자들은 여러 매체들 중에 텔레비전을 통하여 많은 광고를 접하게 됨으로써 궁극적으로 기업의 텔레비전 광고 활동들은 브랜드 이미지에 영향을 미치게 된다. 따라서 본 연구는 국내 텔레비전 광고 표현 요소 중 영상 표현 타입을 중심으로 소비자들이 선호하는 브랜드 이미지가 무엇인가에 관하여 조사하고, 국내 텔레비전 광고 속 제품의 어떤 점이 브랜드 이미지를 형성하는데 중요한 요소로 작용하는 지에 대해 알아보고자 한 연구이다.

• 한국콘텐츠학회논문지
• /
• 제8권9호
• /
• pp.94-100
• /
• 2008

디지털애니메이션은 기본적으로 애니메이션 장르만이 갖는 캐릭터성과 그리고 실사적 영상의 재현구조를 가능하게 하는 기술적 특징을 동시에 가지고 있다. 하지만 그럼에도 불구하고 아직까지 디지털애니메이션으로 제작되어진 극사실주의적 작품들은 관객에게 낯선 존재로 인식되어지고 있다. 최첨단 기술과 인력 그리고 자본으로 탄생된 극사실주의 할리우드 디지털애니메이션에 대한 관객들의 이러한 차가운 반응은 바로 기술적 문제에서라기보다는 미학적 측면에서의 접근법에 그 문제가 존재해 있기 때문이다, 그러므로 본 연구는 디지털애니메이션의 장치적 특성을 기존의 실사영화와의 비교를 통해 기술과 미학에 대한 새로운 접근방법을 제시하고 또한 새로운 리얼리즘으로서 디지털애니메이션의 고유성을 찾아보고자 한다.

• 복식문화연구
• /
• 제29권4호
• /
• pp.603-615
• /
• 2021

The aim of this study was to analyze the semantic network structure of keywords and the visual composition of images extracted from Instagram in relation to the multi-persona phenomenon with in fashion imagery, which has recently been attracting attention. To this end, the concept of a 'secondary character', which forms a separate identity from a 'main character' on various social media platforms as well as on the airwaves, was considered as the spread of multi-persona and #SecondaryCharacter on Instagram was investigated. 3,801 keywords were collected after crawling the data using Python and morphological analysis was undertaken using KoNLP. The semantic network structure was then examined by conducting a CONCOR analysis using UCINET and Netdraw to determine the top 50 keywords. The results were then classified into a total of 6 clusters. In accordance with the meaning and context of the keywords included in each cluster, group names were assigned : virtual characters, relationship with the main character, hobbies, daily record, N-job person, media and marketing. Image analysis considered the technical, compositional, and social styles of the media based on Gillian Rose's visual analysis method. The results determined that Instagram uses fashion images that virtualize one's face to produce multi-persona representation s that show various occupations, describe different types of hobbies, and depict situations pertaining to various social roles.

• 대한전자공학회논문지SP
• /
• 제39권3호
• /
• pp.58-65
• /
• 2002

본 논문에서는 프레임간 압축된 영상의 블록화 현상을 감소시키기 위해 재구성하기 전 차분 영상을 처리하는 새로운 복원 알고리듬을 제안한다. 프레임내 압축 영상의 블록화 현상은 오직 8×8 DCT에 의해서 일어나는 반면에, 프레임간 압축된 영상에서는 8×8 DCT뿐만 아니라 움직임 보상을 위해 사용한 16×16의 매크로 블록에 의해서 발생된다. 이러한 이유로 본 논문에서는 차분 영상에 대한 새로운 열화모델을 제시하고, 블록 경계와 내부의 불연속에 대한 POCS를 이용한 복원 알고리듬을 제시한다. 제안된 알고리듬은 DCT계수의 일부를 이용해 블록 경계의 방향을 고려하는 공간 적응적 저대역 통과 필터를 포함하는 표준 정규화의 변환된 형태이다. 일반적인 비디오 압축 표준은 블록을 기준으로 하는 움직임 보상과 블록 이산여현 변환(BDCT)을 이용한 혼성(hybrid) 구조를 채택하고 있다. 이러한 이유에서 블록화 현상은 블록 경계와 블록 내부에서 일어난다. 두 종류의 블록화 현상을 좀 더 완벽히 제거하기 위해서 복원된 차분 영상은 블록 경계와 블록 내부에서의 방향성 불연속과 같은 제약조건을 만족시켜야만 한다. 그러한 제약조건은 차분 영상을 복구하기 위한 convex set을 정의하는데 이용한다.

• 패션비즈니스
• /
• 제8권3호
• /
• pp.70-83
• /
• 2004

The female image has been changed incessantly with age and has been eminently represented in cosmetic advertisement. The female images are changed from the classical images to the active and professional images with the historical current of cyber, digital, and fusion. These changes in the expression of female image are largely due to the spread of post-modernism, feminism, de-construction (Kim, 1994), digital information revolution, and increase of income. The female images in cosmetic advertisement have been expressed very variously with fashion, marketing target, and characteristics of articles. (Im, 1997) The cosmetic advertisements of pure and graceful images were popular in the past. But nowadays individual image, womanly image, and unchanged beautiful image are in vogue. Individual image is very popular in young generation with very short fashion period. Active career woman image represents passion and beauty with extension of women' social roles. Unchanged beautiful image in modern industrial pollution stands for the desire of keeping the beauty in youth. Brand is very important factor to consumers in purchasing. Brand is no larger the simple concept in the past, and accepted as reflection of the image, social status, service, and life-style(Lee, 1998) consumers are very favorable to foreign cosmetics with the increase of overseas travels and import. This phenomenon is considered as very natural, especially in young generation. To create a successful brand image, the harmony of quality, psychological preference, advertisement, and reasonable price are required. According to the questionnaire research implemented by the subject of college women students in five universities in Seoul, the majority of college women students purchase domestic cosmetic brands in cosmetic specialty stores and depend on the tips of acquaintances. Quality, skin-trouble, and brand are considered as special regards for purchasing. Especially, internet shopping in purchasing marks high growth rate and preference for foreign cosmetics is very ardent. It can be expected that the 21st century is the epoch of various small production different from the mass-production in the 20th century. Female image will be probably expressed with individual, emotional, and professional image in the media of digital, fusion, cyber, and technology culture. It can be said that the tendency of cosmetic purchasing in the future will be more focused on brand image and life style.

• 한국인터넷방송통신학회논문지
• /
• 제21권6호
• /
• pp.155-161
• /
• 2021

최근, 실시간으로 카메라를 통해 동작을 인식하는 기술의 연구가 많이 진행되고 있다. 기존의 연구들에서는 사람의 관절로부터 특징을 추출하는 개수가 적기 때문에 동작 분류의 정확도가 낮은 한계점들이 있다. 본 논문에서는 이러한 한계점들을 해결하기 위해 움직일 때 변하는 관절의 각도를 특징 추출하여 계산하는 알고리즘과 이미지 분류 시에 정확도가 높은 CBAM(Convolutional Block Attention Module)을 사용한 분류모델을 제안한다. AI Hub에서 제공하는 피트니스 자세 이미지로부터 5가지 운동 동작 이미지를 인용하여 분류 모델에 적용한다. 구글에서 제공하는 그래프 기반 프레임워크인 MediaPipe 기법을 사용하여, 이미지로부터 운동 동작 분류에 중요한 8가지 관절 각도 정보를 추가적으로 추출한다. 추출한 특징들을 모델의 입력으로 설정하여, 분류 모델을 학습시킨다. 시뮬레이션 결과로부터 제안한 모델은 높은 정확도로 운동 동작을 구분하는 것을 확인할 수 있다.