• KIPS Transactions on Software and Data Engineering
• /
• v.10 no.11
• /
• pp.457-464
• /
• 2021

The advent of the big data era has enabled the rapid development of deep learning that learns rules by itself from data. In particular, the performance of CNN algorithms has reached the level of self-adjusting the source data itself. However, the existing image processing method only deals with the image data itself, and does not sufficiently consider the heterogeneous environment in which the image is generated. Images generated in a heterogeneous environment may have the same information, but their features may be expressed differently depending on the photographing environment. This means that not only the different environmental information of each image but also the same information are represented by different features, which may degrade the performance of the image analysis model. Therefore, in this paper, we propose a method to improve the performance of the image color constancy model based on Adversarial Learning that uses image data generated in a heterogeneous environment simultaneously. Specifically, the proposed methodology operates with the interaction of the 'Domain Discriminator' that predicts the environment in which the image was taken and the 'Illumination Estimator' that predicts the lighting value. As a result of conducting an experiment on 7,022 images taken in heterogeneous environments to evaluate the performance of the proposed methodology, the proposed methodology showed superior performance in terms of Angular Error compared to the existing methods.

• KSCE Journal of Civil and Environmental Engineering Research
• /
• v.39 no.1
• /
• pp.227-237
• /
• 2019

In terms of the concept of space in urban planning and design, the modern city approaches from the perspective that all spaces are made up of different places based on identity. In this study, we analyzed disparate image by paying attention to the difference between the individual spaces that form the city. And the physical and behavioral factors that respond to the disparate spatial images are derived through legibility measurement. As a result, the urban space in Busan has disparate images such as closure, openness, dynamic + closure, dynamic + openness, quiet space. The following are the legibility factors that make each image stronger. The elements of closure are high-density and high-rise buildings, residence, work, openness appeared as a factor of water space, walking and viewing. Dynamic & Closure is the element of industrial activity, dynamic & openness is open space, water space, swimming element, quiet space image is green space, walking, stay factor.

• Journal of KIISE:Databases
• /
• v.30 no.5
• /
• pp.464-475
• /
• 2003

few sample objects and compressed histogram information of image databases. The histogram information is used to estimate the selectivity of spherical range queries and a small number of sample objects is used to compensate the selectivity error due to the difference of the similarity measures between meta server and local image databases. An extensive experiment on a large number of image data demonstrates that our proposed method performs well in the distributed heterogeneous environment.

• KSCE Journal of Civil and Environmental Engineering Research
• /
• v.37 no.6
• /
• pp.1077-1085
• /
• 2017

Urban space is a place that shows the behavior, experience and lifestyle of residents. Also by the combination of local attributes and assets, creating a space that reflects the unique image of the city. For the development of the city, it is important to actively reflect the identity of the space created by the difference of each space. In this study, we focused on the disparate of space and found the characteristics of urban space. And psychological factors on disparate images of the spatial analysis. As a result, 14 spatial image characteristics of Busan showed difference. In terms of dimensions, it could be classified as an natural, artificial, dynamic, static images. Also, the preference cluster of individuals showed a positive response in naturalness, openness, gentle slope and static space.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2019.05a
• /
• pp.644-646
• /
• 2019

테크놀로지의 발달로 영상 이미지는 영상 예술 뿐 만 아니라 다매체에서 디지털이미지로 전달되고 있다. 기존의 영상 이미지는 화면 프레임을 단순히 시간적 움직임을 재현하는 것이나 제한된 화면 효과를 통해 표현되었다. 그러나 테크놀로지의 진보로 인하여 디지털 미디어는 기존 영상 기법을 확장시켜 다양한 표현을 빠르고 쉽고 편리하게 구현이 가능하게 되었다. 이와 같은 기술적 진화는 결과적으로 영상 이미지가 생성하는 이미지와 이미지 사이에 생기는 이질적 간격을 복잡하고 자극적으로 끝임 없이 전달 받게 되는 상황에 놓이게 되었다. 우리에게 디지털 이미지가 생성하는 이미지의 시간성이 움직임을 어떻게 해석할 수 있는지 들뢰즈의 관점에서 살펴보고, 인식적 요소가 실질적으로 어떻게 표현될 수 있는지 실증적인 사례<아디다스 광고 영상>을 중심으로 분석해 보고자 한다.

• Journal of Advanced Navigation Technology
• /
• v.13 no.2
• /
• pp.257-261
• /
• 2009

This paper propose a new method using Gestalt theory to extract image contour. The proposed method use Gestalt theory based on proximity, similarity and continuation for grouping objects from image segmentation. It use downward feedback and perception to materialize one visual level in image from heterogeneity visual levels in image. The experimental result show that the proposed method achieves better performance than other methods.

• Annual Conference on Human and Language Technology
• /
• 2020.10a
• /
• pp.468-472
• /
• 2020

시각적 질의응답(Visual Question Answering, VQA)은 주어진 이미지에 연관된 다양한 질문에 대한 올바른 답변을 예측하는 기술이다. 해당 기술은 컴퓨터 비전-자연어 처리 연구분야에서 활발히 연구가 진행되고 있으며, 질문의 의도를 정확히 파악하고, 주어진 이미지에서 관련 단서 정보를 찾는 것이 중요하다. 또한, 서로 이질적인 특성을 지닌 정보(이미지 객체, 객체 위치, 질문)를 통합하는 과정도 중요하다. 본 논문은 질문의 의도에 알맞은 정보를 효율적으로 사용하기 위해 멀티 모달 입력 이미지 객체, 객체 위치, 질문)에 대한 Multi-modal Message Aggregation (MMA) 제안하며 이를 통해 한국어 시각적 질의응답 KVQA에서 다른 모델보다 더 좋은 성능을 확인하였다.

• Proceedings of the Korean Information Science Society Conference
• /
• 2000.10a
• /
• pp.48-50
• /
• 2000

최근 텍스트, 이미지, 오디오, 비디오와 같은 멀티미디어 객체를 다루는 연구는 국내외적으로 활발하게 진행되고 있으나, 이러한 멀티미디어 객체들을 효율적으로 저장 및 검색하기 위한 하부저장 시스템에 대한 연구는 미흡한 실정이다. 본 연구에서는 이러한 대용량 멀티미디어 객체들을 효율적으로 저장 및 검색하기 위한 구조를 분석하고 다양한 이질적 객체들을 위한 객체 관리자 및 대용량 텍스트를 위한 역화일 관리자를 설계한다. 아울러, 기존의 하부저정 구조인 SHORE 저장시스템에 통합하여 DBMS 측면에서 제공하는 동시성 제어, 회복기법 등을 지원할 수 있는 객체저장 엔진을 구현한다.

• Cartoon and Animation Studies
• /
• s.49
• /
• pp.203-220
• /
• 2017

In the field of visual arts, the introduction of synthesis allowed us to express various expressions that were outside the constraints of time and space. Digital synthesis is constrained by realistic representation due to technical limitations. However, it is no exaggeration to say that after digital synthesis has been fully established, at least the limits of image representation through synthesis have almost disappeared. The existing research papers on composing are either technical studies on the production techniques, or the synthesis was mainly focused on the change of the space-time concept and meaning of the visual arts. I felt the need for research on. In addition, I felt the need to look back on the changes in my view of art. Therefore, in this paper, rather than dealing with the conceptual theory of the technical part or spatio-temporal extension, it is necessary to classify it as natural, heterogeneous, heterogeneous and natural according to the traits revealed in the artistic expression of art, And diversity.

• Journal of Korea Multimedia Society
• /
• v.9 no.7
• /
• pp.782-800
• /
• 2006

Image metadata is used to describe the contents, structures, and features of a given image data. There exist lots of image metadata-related standards. Such image metadata need to be shared to efficiently search the correct images in Web Service-enabled distributed environments. In this thesis, we propose an interoperable system architecture for utilizing image metadata and query extension methods using metadata mapping for correct and efficient image searching. In the proposed methods, the provider schema using non-standard image metadata are mapped into the schema based on the VRA standard, which is used as a basic and intermediary standard in MDR(MetaData Registry). We support mapping operations between the VRA standard and other standards, to reduce the number of unnecessary mapping operations and to enable easier addition of schema based on new standards. Using the proposed query processing methods, which utilize image metadata and overcome their heterogeneity by metadata mapping, it is possible to severely improve the correctness of image retrieval. To show the usefulness of our schemes, we implemented a prototype system and performed retrieval performance evaluation on extended query processing.