http://chineseinput.net/에서 pinyin(병음)방식으로 중국어를 변환할 수 있습니다.
변환된 중국어를 복사하여 사용하시면 됩니다.
Development of Content-Based Trademark Retrieval System on the World Wide Web
Kim, Young-Sum,Kim, Yong-Sung,Kim, Whoi-Yul,Kim, Myung-Joon Electronics and Telecommunications Research Instit 1999 ETRI Journal Vol.21 No.1
In this paper, we describe a new trademark retrieval system based upon the content or the shape of trademark. The system has an on-line graphical user interface for the World Wide Web (WWW) that allows user to provide a query in forms of a sketch or a visual image to search for similar trademarks from database. User interfaces for the WWW were implemented by utilizing HTML and Java applets. The query can occur in arbitrary size and orientation. A shape representation scheme invariant to scale and rotation was developed to measure the similarity between two trademarks using the magnitude of Zernike moments as a feature set. Performance evaluation has been carried out with a database of 3,000 trademarks. It takes only about 0.6 second for the retrieval on a 200 MHz Pentium PC. The average recall of the original one among top 30 candidates queried by noisy or deformed images was 100%.
Whoi-Yul Kim 한국정보과학회 1997 Journal of Electrical Engineering and Information Vol.2 No.6
In order to achieve scale- and rotation-invariance in recognizing unoccluded objects in binary images using Zemike moment features, an image of an object has often been normalized first by its zeroth-order moment (ZOM) or area. With elongated objects such as characters, a stroke width varies with the threshold value used; it becomes one or two pixels wider or thinner. The variations of the total area of the character becomes significant when the character is relatively thin with respect to its overall size, and the resulting normalized moment features are no longer reliable. This dilation/erosion effect is more severe when the object is not focused precisely. In this paper, we analyze the ZOM method and propose as a normalization method, the maximum enclosing circle (MEC) centered at the centroid of the character. We compare both the ZOM and MEC methods in their performance through various experiments.
Kim, Whoi-Yul The Korean Institute of Electrical Engineers 1997 Journal of Electrical Engineering and Information Vol.2 No.6
In order to achieve scale- and rotation-invariance in recognizing unoccluded objects in binary images using Zernike moment features, an image of an object has often been normalized first by its zeroth-order moment (ZOM) or area. With elongated objects such as characters, a stroke width varies with the threshold value used, it becomes one or two pixels wider or thinner. The variations of the total area of the character becomes significant when the character is relatively thin with respect to its overall size, and the resulting normalized moment features are no longer reliable. This dilation/erosion effect is more severe when the object is not focused precisely. In this paper, we analyze the ZOM method and propose as a normalization method, the maximum enclosing circle (MEC) centered at the centroid of the character. We compare both the ZOM and MEC methods in their performance through various experiments.
Fast Computation of Zernike Moments Using Three Look-up Tables
Kim, Sun-Gi,Kim, Whoi-Yul,Kim, Young-Sum,Park, Chee-Hang The Korean Institute of Electrical Engineers 1997 Journal of Electrical Engineering and Information Vol.2 No.6
Zernike moments have been one of the most commonly used feature vectors for recognizing rotated patterns due to its rotation invariant characteristics. In order to reduce its expensive computational cost, several methods have been proposed to lower the complexity. One of the methods proposed by mukundan and K. R. Ramakrishnan[1], however, is not rotation invariant. In this paper, we propose another method that not only reduces the computational cost but preserves the rotation invariant characteristics. In the experiment, we compare our method with others, in terms of computing time and the accuracy of moment feature at different rotational angle of an object in image.
전역 실루엣 및 지역 광류 특징을 이용한 사람의 동작 인식
김현철(Kim, HyunCheol),나문수(Ra, Moon-Soo),김희권(Kim, Hee-Kwon),남승우(Nam, Seung-Woo),이재호(Lee, Jae-Ho),김회율(Kim, Whoi-Yul) 한국방송·미디어공학회 2011 한국방송공학회 학술발표대회 논문집 Vol.2011 No.11
인간의 동작 인식은 가상 현실 시스템 및 게임 등에 적용할 수 있는 컴퓨터 비전 분야의 요소 기술 중 하나로써, 최근까지 그 연구과 활발히 진행되고 있다. 본 논문에서는 빠르고 정확한 동작 인식을 위해, 실루엣과 모션 특징이 결합된 방법을 제안한다. 제안하는 방법은 전역 특징을 이용한 후보 동작 선정 및 지역 특징을 이용한 검증 2 단계로 구성된다. 전역 특징은 Motion History Image의 Hu 모멘트를 이용해 계산되며, 후보 동작의 선정은 이들의 통계치를 이용해 결정한다. 한정된 후보 동작들 중 입력 동작을 정확히 인식하기 위해, 공간 및 방향성 비닝 기법으로 추출된 광류와 실루엣 특징을 지역 특징으로 이용한다. 최종 인식 결과는 Hu 모멘트 통계치와의 유사도 및 지역 특징의 학습을 통해 생성된 Support Vector Machine의 결과를 고려하여 결정된다. 제안하는 방법의 성능을 평가하기 위해, 실세계에서 사용 빈도가 높으며 동작의 변화가 큰 13 개의 제스처를 선정하여 데이터 셋을 구성하였다. 실험 결과 제안하는 방법의 연산 시간은 50 ms, 인식 정확도는 95%임을 확인하였다.
동적 배경에 강인한 Propagation 기반의 코드북 배경 모델링 방법
김도영(Doyung Kim),김중식(Joongsik Kim),김회율(Whoi-Yul Kim) 대한전자공학회 2015 대한전자공학회 학술대회 Vol.2015 No.6
In video surveillance system, the background modeling techniques for detecting moving objects such as pedestrians and vehicles are essential to separate foreground and background. However, extracting foreground in changing environmental condition is difficult because of the illumination change, moving background and noise. In this paper, we proposed the background modeling technique based on Codebook using the propagati on of adjacent pixel. This paper compares the results of other algorithms.
Fast Computation of Zornike Moments Using Three Look-up Tables
Sun-Gi Kim,Whoi-Yul Kim,Young-Sum Kim,Chee-Hang Park 한국정보과학회 1997 Journal of Electrical Engineering and Information Vol.2 No.6
Zemike moments have been one of the most commonly used feature vectors for recognizing rotated patterns due to its rotation invariant characteristics. In order to reduce its expensive computational cost, several methods have been proposed to lower the complexity. One of the methods proposed by Mukundan and K. R. Ramakrishnan [1], however, is not rotation invariant. In this paper, we propose another method that not only reduces the computational cost but preserves the rotation invariant characteristics. In the experiment, we compare our method with others, in terms of computing time and the accuracy of moment feature at different rotational angle of an object in image.
적응적 매개변수 갱신을 통한 효과적인 그림자 제거 기법
김병수(Byeoung-su Kim),이광국(Gwang-Gook Lee),윤자영(Ja-Young Yoon),김재준(Jae-Jun Kim),김회율(Whoi-Yul Kim) 대한전자공학회 2008 電子工學會論文誌-SP (Signal processing) Vol.45 No.3
영상 내에서 이동하는 객체를 추출하는 전경 분리 방법은 객체의 위치 추적 및 인식에 있어서 필수적인 기술이다. 하지만 이동하는 객체 주변에 그림자가 발생하는 경우 이러한 전경 분리 방법에서는 그림자도 전경 영역으로 잘못 판단하여 분리하게 되어 이동 객체의 정확한 형태를 파악하거나 위치를 추정하기 어려운 문제가 있다. 본 논문에서는 이러한 문제를 해결하기 위하여 색상 정보를 이용하여 그림자를 모델링하고 이를 통해 전경 영역 내의 그림자 화소를 Bayesian 분류법에 따라 제거하는 방법을 제안하였다. 특히 제안하는 방법은 매개변수 갱신 과정을 통해 그림자의 특성이 동적으로 모델링되기 때문에 주변 조명의 지속적인 변화에 적응적으로 대응할 수 있다. 실험 결과 제안하는 방법은 다양한 환경에서 그림자를 효과적으로 제거하는 것을 확인하였다. Background subtraction, which separates moving objects in video sequences, is an essential technology for object recognition and tracking. However, background subtraction methods are often confused by shadow regions and this misclassification of shadow regions disturbs further processes to perceive the shapes or exact positions of moving objects. This paper proposes a method for shadow elimination which is based on shadow modeling by color information and Bayesian classification framework. Also, because of dynamic update of modeling parametres, the proposed method is able to correspond adaptively to illumination changes. Experimental results proved that the proposed method can eliminate shadow regions effectively even for circumstances with varying lighting condition.