• Annual Conference of KIPS
• /
• 2004.05a
• /
• pp.35-38
• /
• 2004

기존의 내용기반 이미지 검색 시스템은 색상, 질감, 모양등과 같은 특징 벡터를 추출하여 검색하는 방법이 많이 연구되어 왔다. 특히 색상 정보는 이미지를 검색하기 위하여 중요한 정보로 사용되고 있다. 따라서 색상 이미지를 검색하기 위해서 평균 RGB, HSI값을 이용하거나 히스토그램을 이용하는 방식이 많이 사용 되어왔다. 본 논문에서는 사람이 시각적으로 보고 느끼는 색상(H), 채도(S), 명도(I) 방식을 이용한 HSI값을 사용하여 색상 의미를 이용한 지능적 이미지 검색 기법을 제안하고 알고리즘을 설명한다. 색상 의미(Color Semantics)란 사람의 시각적인 특징을 기반으로 칼라 이미지에 적용하여 감성 형용사 기반으로 검색할 수 있는 방법이다. 색상 의미를 이용한 지능적 이미지 검색은 색상-기반 질의(color-based retrieval)를 제공할 뿐만 아니라 인간의 감성이나 느낌에 의한 의미-기반 질의(semantic-based retrieval)방식을 가능하게 한다. 즉, "시원한 이미지" 혹은 "부드러운 이미지"를 검색하는 방식이다. 따라서 사용자의 검색 의도를 보다 정확하게 표현할 수 있으며, 검색의 결과에 대한 만족도를 향상 시킬 수 있다.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2019.05a
• /
• pp.125-128
• /
• 2019

이미지를 보호하기 위한 보안 수단으로는 크게 워터마킹, 스테가노그래피 기법 등이 있다. 그 중 워터마킹 기법은 배경 이미지 위에 사용자가 워터마크로 사용하기 위해 원하는 텍스트나 이미지를 새겨 다른 사람이 사용했을 때, 새겨둔 워터마크를 통해 저작권 등을 증명할 수 있는 용도로 사용하여 이미지를 보호하는 기법이며, 워터마크의 속성에 따라 텍스트 워터마킹, 이미지 워터마킹으로 나뉜다. 본 논문에서는 워터마킹 기법을 이용한 이미지 생성 시, 워터마크 대상 텍스트나 이미지의 크기 및 배치 등을 알맞게 설정하는 방법을 제시하였다. 워터마크 대상의 크기를 배경 이미지에 알맞게 자동으로 조절해 주거나 사용자가 임의로 지정하도록 하여 더욱 자유롭고 효과적인 워터마킹 기법을 구현하였다.

• Annual Conference of KIPS
• /
• 2010.04a
• /
• pp.896-897
• /
• 2010

사용자의 참여와 공유형 서비스들이 증가함에 따라, 웹 상에서의 이미지의 종류와 수도 증가하였다. 본 연구에서는 보다 효율적인 이미지 검색을 위하여, 이미지의 특정 요소만을 반영한 분석과 검색 방법이 아닌 다양한 속성을 융합적으로 고려한 방법을 제안한다. 본 논문에서는 하나의 이미지에서 이미지가 가지고 있는 색의 조화, 주요 색상, 형태, 느낌, 질감 다섯 가지 속성을 이미지 특성 추출 기술을 통해 추출하고, 이를 각각의 속성에 대하여 폴리곤으로 나타내는 이미지 속성 그래프를 생성하는 방안을 제안한다. 또한, 질의에 따라 속성의 우선순위를 부여하고, 폴리곤 매칭을 통하여 사용자의 목적에 맞게 이미지를 분석하고 정보를 제공하는 방법을 설명한다.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2021.10a
• /
• pp.50-51
• /
• 2021

In this paper, in order to design an image deep learning algorithm using a Lunar New Year image, a preliminary study on the shape and shadow of the image is conducted. In order to perform image deep learning, it is necessary to identify the characteristics of the Lunar New Year image, configure an appropriate label, and proceed with the preprocessing process. Image data is a cohort photo collected by Daejeon University, and based on this, we intend to establish a goal for conducting research from the data.

• Proceedings of the Korean Society of Computer Information Conference
• /
• 2023.07a
• /
• pp.573-574
• /
• 2023

화염 영역 검출을 위해 이전 기법에서는 화재 이미지에서 연기제거 및 색상보정을 통해 이미지를 전처리하였다. 그러나 이 기법은 임계값에 영향을 많이 받고, 밝기채널을 이용하여 검출하기 때문에 밤에 일어난 화재 이미지에서는 평균이상의 퍼포먼스를 수행하지만, 주변이 밝은 대낮의 화재 이미지에서는 퍼포먼스가 줄어드는 문제가 있다. 이를 보완하고자 본 논문에서는 CycleGAN을 이용하여 낮 이미지를 밤 이미지로 바꾸어 이미지 전처리를 진행하는 기법을 제안함으로써 화염 감지의 정확도가 개선되었음을 실험을 통해 보여준다.

• Journal of Korea Society of Industrial Information Systems
• /
• v.17 no.3
• /
• pp.83-94
• /
• 2012

This study was designed to identify and develop the measurement items of symbolic image (personality) of most preferred Korean companies in adaptation of prior research of brand personality. From the common factor analysis, factors were extracted to five dimensions: Sincerity, Competence, Affection, Excitement, and Sophistication. In this survey, we collected data from the 558 respondents and performed a structural equation model using AMOS 16.0. The result of the confirmatory factor analysis indicated that the measurement model of Korean corporate symbolic image could be effectively used to envision the personality of Korean companies with similar context of brand personality research. Furthermore, this empirical study showed interesting results. One of unique Korean emotions, affection factor, was derived from this research as well and it is also proven that the Korean-corporate symbolic image could be used to measure the corporate personality and it should be continuously reinterpreted with deviation of socio-cultural changes rather than adhering to standardized paradigm.

• Archives of design research
• /
• v.19 no.4 s.66
• /
• pp.31-40
• /
• 2006

For a long time, we have been good at the static image, 'the country of calm morning' as reminding Korea. But, it became a chance that we were recognized newly and actively through the success of the Olympic games in 1988 and World cup games in 2002 all over the world. And foreigners' interest and expectation for Korea becomes much higher by these reasons. Moreover, the importance of our cultural industry, recognition for new cultural improvement have been changing step by step by the oriental trend in Asia in these days. So, this study intends to realize the Korean symbol with showing Korean Identity and its image's applying to the street furniture design through the making of special street which expresses well the Korean Image. There is a purpose of this study not to use as just street furniture but to make characterized Korean culture through showing examples which are based on Korean culture. The way of the improvement for the street furniture design using Korean image are these we analyze the character of Korean international image with the previous data for the change of Korean Image. We understand the symbol of changed Korean international Image, and extract the design elements. We analogize the shape, material and color image of the street furniture which is joined with both the Korean national image and formative traditional elements. At last, we formalize its symbolic image by applying to the street furniture according to the design improvement process with the analogized image. The formal study using Korean image makes more cleary the Korean Identity, and it doesn't influence only the actual change but also Korean cultural value equipment.

• Journal of the Korea Society of Computer and Information
• /
• v.25 no.9
• /
• pp.63-69
• /
• 2020

Image-image translation is a technology that creates a target image through input images, and has recently shown high performance in creating a more realistic image by utilizing GAN, which is a non-map learning structure. Therefore, there are various studies on image-to-image translation using GAN. At this point, most image-to-image translations basically target one attribute translation. But the data used and obtainable in real life consist of a variety of features that are hard to explain with one feature. Therefore, if you aim to change multiple attributes that can divide the image creation process by attributes to take advantage of the various attributes, you will be able to play a better role in image-to-image translation. In this paper, we propose Multi CycleGAN, a dual attribute transformation structure, by utilizing CycleGAN, which showed high performance among image-image translation structures using GAN. This structure implements a dual transformation structure in which three domains conduct two-way learning to learn about the two properties of an input domain. Experiments have shown that images through the new structure maintain the properties of the input area and show high performance with the target properties applied. Using this structure, it is possible to create more diverse images in the future, so we can expect to utilize image generation in more diverse areas.

• The Journal of the Korea Contents Association
• /
• v.18 no.12
• /
• pp.118-129
• /
• 2018

In AI chatbot service via a messenger, a profile image of the chatbot is the first thing that users see to communicate with the chatbot. This profile image not only manages an impression about the profile owner in SNS on followers, but also makes an important impression about chatbot services on users. Thus motivated, this study investigates proper profile images tailored for the types of chatbot services and users. Specifically, I reviewed the preferred images and expressions of chatbots for each purpose of chatbot service. Then, in a case study, I collected and analyzed the representative chatbot profile images for the purpose of fun and counseling. The profile images are categorized as robot, human, animal, and abstract images. Based on these categories, I surveyed the preferred profile image of the chatbot service in either the text type or image type alternatives. For the purpose of fun, in the text version, I found that both men and women preferred a human image to others. However, in the image version, men preferred woman and robot images while women preferred cute animation character and robot images. For counseling services, both men and women preferred woman and animal images most, which is similar to the results of the text version of questionnaires as well. While both genders consistently preferred real photo images, women tend to like abstract images more than men do. I expect that the results of this study would be useful to develop the proper profile images of AI chatbot for each service purpose.

• Convergence Security Journal
• /
• v.18 no.3
• /
• pp.53-59
• /
• 2018

In this paper, we propose a hierarchical image classification scheme for efficient object image classification. In the non-hierarchical image classification, which classifies the existing whole images at one time, it showed that objects with relatively similar shapes are not recognized efficiently. Therefore, in this paper, we introduce the image classification method in the hierarchical structure which attempts to classify object images hierarchically. Also, we introduce to the efficient class structure and algorithms considering the scalability that can occur when a deep learning image classification is applied to an actual system. Such a scheme makes it possible to classify images with a higher degree of confidence in object images having relatively similar shapes.