• 한국군사과학기술학회지
• /
• 제15권3호
• /
• pp.298-305
• /
• 2012

In this paper, we propose an integrated algorithm for scene recognition, which has been a challenging computer vision problem, with application to mobile robot localization. The proposed scene recognition method utilizes SIFT and visual words as local-level features and GIST as a global-level feature. As local-level and global-level features complement each other, it results in improved performance for scene recognition. This improved algorithm is of low computational complexity and robust to image distortions.

• International journal of advanced smart convergence
• /
• 제9권4호
• /
• pp.101-105
• /
• 2020

In this study, we are going to discuss the role of sound, which exists as the background of a movie scene, and the relationship between seeing and hearing. In order to analyze the form in which the scene and sound are matched, the visual content of the image must be analyzed first. In this study, we analyzed Stanley Kubrick's film . Generally, there are both visual and auditory features in the scene of the movie. In some cases, the sound is emphasized more than the scene, and the scene is emphasized more than the sound. In particular, the sound in movies enables a wider imagination and a three-dimensional experience. Images and sounds will have different meanings and effects depending on the purpose of use. By analyzing the correlation between visual scene and auditory sound based on the role sound plays in the film in this study, we would like to confirm that sound is one essential element that has the role of equivalent and independent expression, not the role of assisting images.

• 한국실내디자인학회:학술대회논문집
• /
• 한국실내디자인학회 2007년도 춘계학술대회 논문집
• /
• pp.176-180
• /
• 2007

The stage space in theatre for the performance of a play requires two aspects of the physical space for a play and the image as the background or the theme of a play. This research focuses on the scene design that creates the background or the theme of a play and communicates it to the audience. The scene design is composed of various design elements of space and objects as the properties of a play. Design elements and the objects are the design languages in various form; plane, three-dimensional, multi-dimensional form. These design language have the significant meaning as a sign like human language. The scene design is completed with rhetorical expression for communication in theatre. This research defines the category of meaning that design elements of scene design can create and the rhetorical expression of the scene design language. King Lear directed by Robert G. Anderson was analyzed with the category of design elements as a sign and the pattern of the rhetorical expression. The scene design for a play is completed effectively by the rhetorical expression of design elements as the design language for communication with the audience in theatre.

• 한국HCI학회:학술대회논문집
• /
• 한국HCI학회 2008년도 학술대회 1부
• /
• pp.708-711
• /
• 2008

Time-of-flight(ToF) cameras deliver intensity data as well as range information of the objects of the scene. However, systematic problems during the acquisition lead to distorted values in both distance and amplitude. In this paper we propose a method to acquire reliable distance information over the entire scene correcting each information based on the other data. The amplitude image is enhanced based on the depth values and this leads depth correction especially for far pixels.

• 대한전자공학회논문지
• /
• 제26권8호
• /
• pp.1247-1256
• /
• 1989

In this paper, we propose the dynamic scene segmentation algorithm using a cross mask and edge information. This method, a combination of the conventioanl feature-based and pixel-based approaches, uses edges as features and determines moving pixels, with a cross mask centered on each edge pixel, by computing similarity measure between two consecutive image frames. With simple calcualtion the proposed method works well for image consisting of complex background or several moving objects. Also this method works satisfactorily in case of rotaitional motion.

• 전자공학회논문지S
• /
• 제36S권10호
• /
• pp.94-102
• /
• 1999

본 논문에서는 카메라의 조명궤적과 화상내의 광휘점 (highlight) 영역의 화소들이 형성하는 색도좌표 군집에 대한 근사된 직선과의 교점을 카메라 촬영 장면에서의 조명색으로 결정하는, 카메라 촬영 장면의 조명색을 화상으로부터 직접 추정하는 방법을 제안하였다. 이 방법은 카메라의 조명에 대한 반응 특성을 사전 정보로 제공함으로써 기존의 화상내의 광휘점 정보망을 이용하던 방법에 비해 더욱 정확한 조명색 추정을 가능하게 한다.

• 전자공학회논문지
• /
• 제50권4호
• /
• pp.163-170
• /
• 2013

일반적으로 외부에서 획득되는 영상은 대기 중에 존재하는 먼지, 물방울, 연무, 안개, 연기 등에 의해 화질이 감쇠되고 결과적으로 대비도 감소와 색상의 왜곡 현상이 발생한다. 그리나 안개와 배경 사이에 내재된 모호성 때문에 배경으로부터 안개를 제거하는 작업은 결코 간단한 문제가 아니다. 본 논문에서는 단일 영상에서 비용함수로서 에지의 기울기를 이용한 그래프 기반 영역 분할 방법을 이용하여 안개 제거를 위한 새로운 방법을 제안한다. 우리는 장면을 깊이 관련 정보에 따라 여러 영역으로 분리하고 전역적인 안개값을 추정한다. 매체의 전달량은 그래프 기반 영역 분할 알고리즘의 임계 함수에 의해서 직접적으로 계산된다. 매체 전달량과 안개값이 계산되면 안개 모델식에 의해서 쉽게 안개가 제거된 영상을 복원할 수 있다. 그리고 안개 영상과 복원된 영상간의 에지의 기울기 비율을 계산함으로써 기존의 연구 방법과 제안된 연구 방법의 가시성 복원 정도를 비교 평가하였다. 다양한 안개 영상에 대한 실험 결과 제안된 방법의 우수한 안개 제거 및 화질 복원 능력이 입증되었다.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제16권7호
• /
• pp.2390-2406
• /
• 2022

Scene text recognition has important application value and attracted the interest of plenty of researchers. At present, many methods have achieved good results, but most of the existing approaches attempt to improve the performance of scene text recognition from the image level. They have a good effect on reading regular scene texts. However, there are still many obstacles to recognizing text on low-quality images such as curved, occlusion, and blur. This exacerbates the difficulty of feature extraction because the image quality is uneven. In addition, the results of model testing are highly dependent on training data, so there is still room for improvement in scene text recognition methods. In this work, we present a natural scene text recognizer to improve the recognition performance from the feature level, which contains feature representation and feature enhancement. In terms of feature representation, we propose an efficient feature extractor combined with Representative Batch Normalization and ResNet. It reduces the dependence of the model on training data and improves the feature representation ability of different instances. In terms of feature enhancement, we use a feature enhancement network to expand the receptive field of feature maps, so that feature maps contain rich feature information. Enhanced feature representation capability helps to improve the recognition performance of the model. We conducted experiments on 7 benchmarks, which shows that this method is highly competitive in recognizing both regular and irregular texts. The method achieved top1 recognition accuracy on four benchmarks of IC03, IC13, IC15, and SVTP.

• 대한원격탐사학회:학술대회논문집
• /
• 대한원격탐사학회 2008년도 International Symposium on Remote Sensing
• /
• pp.137-140
• /
• 2008

Recently many areas require wide field of view images. Such as surveillance, virtual reality, navigation and 3D scene reconstruction. Conventional camera systems have a limited filed of view and provide partial information about the scene. however, omni directional vision system can overcome these disadvantages. Acquiring 360 degree panorama images requires expensive omni camera lens. In this study, 360 degree panorama image was tested using a low cost optical reflector which captures 360 degree panoramic views with single shot. This 360 degree panorama image system can be used with detailed positional information from GPS/INS. Through this study result, we show 360 degree panorama image is very effective tool for mobile monitoring system.

• 한국콘텐츠학회논문지
• /
• 제12권12호
• /
• pp.1-13
• /
• 2012

3차원 동영상은 다양한 응용분야들에서 차세대 콘텐츠로 큰 주목을 받고 있다. 2D-to-3D 변환은 3차원 동영상의 시대로 넘어가는 과도기 동안에 3차원 동영상 콘텐츠의 부족현상을 해결하기위한 강력한 기술로 여겨지고 있다. 일반적으로 2D-to-3D 변환을 위해서는 2차원 동영상 각 장면의 깊이영상을 추정/생성한 후 깊이 영상 기반 랜더링 (DIBR : Depth Image Based Rendering) 기술을 이용하여 스테레오 동영상을 합성한다. 본 논문은 2차원 동영상 내 존재하는 다양한 변환 단서들을 통합하는 새로운 깊이 융합 기법을 제안한다. 우선, 알맞은 깊이 융합을 위해 몇몇 단서가 현재 장면을 효과적으로 표현할 수 있는 지 아닌지 검사된다. 그 후, 신뢰성 검사의 결과를 기반으로 현재 장면은 4개의 유형 중 하나로 분류된다. 마지막으로 최종 깊이 영상을 생성하기 위해 신뢰할 수 있는 깊이 단서들을 조합하는 장면 적응적 깊이 융합이 수행된다. 실험 결과를 통해 각각의 단서가 장면 유형에 따라 타당하게 활용되었고 최종 깊이 영상이 현재 장면을 효과적으로 표현할 수 있는 단서들에 의해 생성되었음을 관찰할 수 있다.