• 한국안전학회지
• /
• 제12권3호
• /
• pp.132-138
• /
• 1997

For many years concrete has been the major building material for most construction. It is of primary importance that fire fighters or fire investigators have a full understanding of the properties of concrete so that better control of the fire scene is achieved. This, in turn, not only help to ensure a safer fire-fighting job but also a more successful fire investigation. So far as the fire scene investigation in concerned, knowledge about the thermal behaviour of concrete can help the investigators to determine the highest temperature that a particular spot of a fire scene has ever reached thereby providing data which may be of value in reconstructing the course of the fire.

• Journal of Information Processing Systems
• /
• 제11권4호
• /
• pp.528-537
• /
• 2015

Text in images is one of the most important cues for understanding a scene. In this paper, we propose a novel approach based on interest points to localize text in natural scene images. The main ideas of this approach are as follows: first we used interest point detection techniques, which extract the corner points of characters and center points of edge connected components, to select candidate regions. Second, these candidate regions were verified by using tensor voting, which is capable of extracting perceptual structures from noisy data. Finally, area, orientation, and aspect ratio were used to filter out non-text regions. The proposed method was tested on the ICDAR 2003 dataset and images of wine labels. The experiment results show the validity of this approach.

• 만화애니메이션 연구
• /
• 통권9호
• /
• pp.199-219
• /
• 2005

본 논문 에서는 영상매체의 몽타주 이론과 미장센의 효과 등을 살펴보고, 구체적인 작품의 사례들과 그것이 애니메이션에 어떠한 효과와 발전에 도움이 되었나를 살펴보고자 한다. 아울러 이러한 시각매체의 대표적인 장르인 영화의 영상연구를 바탕으로 몽타주이론과 미장센이 애니메이션의 장면에 도입되어 나타나는 점들도 고찰해 보겠다. 향후 애니메이션의 발전에 영상의 몽타주 이론과 미장센의 폭넓은 이해와 수용 발전으로 창의적인 애니메이션의 장면들에 도움을 주고자 하며, 더 나아가 영상예술로서의 애니메이션의 독특하고 창조적인 화면을 위해 미장센과 몽타주 이론의 폭넓은 수용을 통해 애니메이션의 발전에 이바지 하려는 연구 목적이 있다.

• 전자통신동향분석
• /
• 제38권6호
• /
• pp.12-21
• /
• 2023

Visual relationship understanding in computer vision allows to recognize meaningful relationships between objects in a scene. This technology enables the extraction of representative information within visual content. We discuss the technology of visual relationship understanding, specifically focusing on videos. We first introduce visual relationship understanding concepts in videos and then explore the latest existing techniques. Next, we present benchmark datasets commonly used in video visual relationship understanding. Finally, we discuss future research directions in video visual relationship understanding.

• ETRI Journal
• /
• 제42권3호
• /
• pp.399-410
• /
• 2020

Detecting visual relationships in an image is important in an image understanding task. It enables higher image understanding tasks, that is, predicting the next scene and understanding what occurs in an image. A visual relationship comprises of a subject, a predicate, and an object, and is related to visual, language, and spatial cues. The predicate explains the relationship between the subject and object and can be categorized into different categories such as prepositions and verbs. A large visual gap exists although the visual relationship is included in the same predicate. This study improves upon a previous study (that uses language cues using two losses) and a spatial cue (that only includes individual information) by adding relative information on the subject and object of the extant study. The architectural limitation is demonstrated and is overcome to detect all zero-shot visual relationships. A new problem is discovered, and an explanation of how it decreases performance is provided. The experiment is conducted on the VRD and VG datasets and a significant improvement over previous results is obtained.

• 한국정보과학회논문지:소프트웨어및응용
• /
• 제34권6호
• /
• pp.549-561
• /
• 2007

영상을 분석하여 얻은 증거를 바탕으로 장면의 의미를 추론하고 해석하는 것을 시각 기반 장면 이해라고 하며, 최근 인과적인 판단 및 추론 과정을 모델링하기에 유리한 베이지안 네트워크(BN)를 이용한 확률적인 접근 방법이 활발히 연구되고 있다. 하지만 실제 환경은 변화가 많고 불확실하기 때문에 의미 있는 증거를 충분히 확보하기 어려울 뿐만 아니라 전문가에 의한 설계로 유지하기 어렵다. 본 논문에서는 증거 및 학습 데이타가 부족한 장면인식 문제에서 효율적인BN 구조로 계산 복잡도가 줄어들고 정확도는 향상될 수 있는 BN 학습방법을 제안한다. 이 방법은 추론 대상 환경의 도메인 지식을 온톨로지로 표현하고 이를 제한적으로 사용하여 효율적인 계층구조의 BN을 구성한다. 제안하는 방법의 평가를 위하여 9종류의 환경에서 90장의 영상을 수집하고 레이블링하여 실험하였다. 실험 결과, 제안하는 방법은 증거의 수가 적은 불확실한 환경에서도 좋은 성능을 내고 학습의 복잡도가 줄어듦을 확인할 수 있었다.

• 제어로봇시스템학회논문지
• /
• 제15권5호
• /
• pp.505-510
• /
• 2009

In recent years the space where a robot works has been expanding to the human space unlike traditional industrial robots that work only at fixed positions apart from humans. A human in the recent situation may be the owner of a robot or the target in a robotic application. This paper deals with the latter case; when a robot vision system is employed to monitor humans for a surveillance application, each person in a scene needs to be identified. Humans, however, often move together, and occlusions between them occur frequently. Although this problem has not been seriously tackled in relevant literature, it brings difficulty into later image analysis steps such as tracking and scene understanding. In this paper, a probabilistic neural network is employed to learn the patterns of the best dividing position along the top pixels of an image region of partly occlude people. As this method uses only shape information from an image, it is simple and can be implemented in real time.

• 반도체디스플레이기술학회지
• /
• 제16권2호
• /
• pp.70-74
• /
• 2017

Given two stereo images of a scene, it is possible to recover a 3D understanding of the scene. This is the primary way that the human visual system estimates depth. This process is useful in applications like robotics, where depth sensors may be expensive but a pair of cameras is relatively cheap. In this work, we combined our interests to implement a graph cut algorithm for stereo correspondence, and performed evaluation against a baseline algorithm using normalized cross correlation across a variety of metrics. Experimental trials revealed that the proposed descriptor exhibited a significant improvement, compared to the other existing methods.

• 한국가시화정보학회:학술대회논문집
• /
• 한국가시화정보학회 2002년도 추계학술대회 논문집
• /
• pp.77-80
• /
• 2002

The recent 3D visualization such as volume rendering, iso-surface rendering or stream line visualization gives more understanding about structures or distribution of data in a space and, moreover, the real-time rendering of a scene enables the animation of time-series data. Because the meteorological data is frequently formed as multi-variables, 3-dimensional and time-series data, the spatial analysis, time-series analysis, vector display, and animation techniques can do important roles to get more understanding about data. In this research, our aim is to develop the 3-dimensional visualization techniques for meteorological data in the PC environment by using IDL. The visualization technology from :his research will be used as basic technology not only for the deeper understanding and the more exact prediction about meteorological environments but also for the scientific and spatial data visualization research in any field from which three-dimensional data comes out such as oceanography, earth science, or aeronautical engineering.

• 인지과학
• /
• 제5권2호
• /
• pp.47-73
• /
• 1994

본 논문에서는 일반적인 도로 영상을 이해할 수 있는 시스템을 블랙보드 모델을 이용하여 구현하였다. 블랙보드에는 계층적인 구조를 갖는 여러 가지의 정보를 저장 하도록 하였으며 이들은 제어모듈의 통제에 따라 여러 개의 지식원들과 유기적으로 결 합하여 가정을 세우고 검증하므로써 도로 영상을 이해하도록 하였다. 실제의 영상을 대상으로 실험한 결과는 90% 정도의 물체를 인식하는 것을 확인하였다. 이 결과를 토 대로 무인운항에 필요한 도로 정보의 추출이 가능하다는 것을 확인하였다.