• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 1999.11a
• /
• pp.416-419
• /
• 1999

We Present a technique for recognizing facial expressions from image sequences. The technique uses muscle-based feature models for tracking facial features. Since the feature models are constructed with a small number of parameters and are deformable in the limited range and directions, each search space for a feature can be limited. The technique estimates muscular contractile degrees for classifying six principal facial express expressions. The contractile vectors are obtained from the deformations of facial muscle models. Similarities are defined between those vectors and representative vectors of principal expressions and are used for determining facial expressions.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2003.05a
• /
• pp.643-646
• /
• 2003

본 논문은 분할된 비디오 화면들의 색상 내용을 기반으로 하는 새로운 영상 검색 방법을 제안 하고자 한다. 제안된 시간 색상 상관그래프는 공통적인 통계 데이터를 이용하여 비디오 화면 내의 공간-시간 관계를 계산한다. 시간 색상 상관 그래프는 내용 기반의 영상 검색에 매우 효과적인 것으로 밝혀진 HSV(Hue, Saturation, Value) 색상 상관 그래프를 기반으로 하고 있다. 시간 색상 상관 그래프는 하나의 비디오 화면으로부터 추출된 프레임 샘플의 양자화 된 HSV 색상 값의 자기상관관계를 이용하여 구성하였다. 본 논문에서는 11시간 분량의 분할된 MPEG-1 비디오에 대한 질의와 질의에 대한 관련성 판정을 하고자 내용 기반의 멀티미디어 검색 시스템을 구축하여 실험하였다. 실험 견과 제안된 방법이 시각 정보만을 필요로 하는 검색에 있어 기존의 다른 검색 방법보다 우수한 결과를 나타냄을 증명하였다.

• Proceedings of the Korea Multimedia Society Conference
• /
• 2004.05a
• /
• pp.203-206
• /
• 2004

본 논문에서는 다양하고 실재적인 조명과 얼굴방향에 관계없이 원활하게 실시간으로 눈을 추적하는 방법을 제안하고자 한다. 기존의 능동적 적외선 조명을 이용한 대다수의 눈 추적장치들은 밝은 동공효과를 이용하고 있다. 그러나, 눈 깜박임, 외부 조명 간섭과 같은 여러 가지 요소로 인하여 동공들이 충분하게 밝게 나타나지 않는 경우가 많이 있다. 그러므로, 본 논문에서는 능동적 적외선 조명을 기반으로 한 칼만 필터링을 이용한 객체 추적 방법과 전형적인 외관을 기반으로 객체 인식 방법을 결합함으로써, 외부 조명의 간섭으로 밝은 동공 효과가 나타나지 않는 경우에도 견실하게 눈을 검출하고 추적 할 수 있는 방법을 제안한다. 눈 검출과 추적을 위해 SVM과 평균 이동 추적 방법을 사용하였고, 적외선 조명과 카메라를 포함한 영상 획득 장치를 구성하여 기존의 방법과 비교 실험한 결과 제안된 방법은 일부 피검자의 경우 100% 완벽하게 눈 추적을 할 수 있음을 보여 주었다.

• Journal of Life Science
• /
• v.21 no.12
• /
• pp.1666-1677
• /
• 2011

The regulation of gene expression plays an important role in cell cycle controls. In this study, a novel gene, the $mas1^+$($\underline{mi}$tosis $\underline{as}$sociated protein) gene, a homolog of human CIP29/Hcc1, was isolated and characterized from fission yeast Schizosaccharomyces pombe (S. pombe) using a gene-specific polymerase chain reaction. The isolated gene contained a complete open reading frame capable of encoding 245 amino acid residues with a typical promoter, as judged by nucleotide sequence analysis. It was also found that a PCB ($\underline{p}$ombe cell $\underline{c}$ycle $\underline{b}$ox) is located in the promoter region, which controls M-$G_1$ specific transcription in S. pombe. The quantitative analysis of the $mas1^+$ transcript against $adh1^+$ showed that the pattern of expression is similar to that of the septation index. Cytokinesis of mas1 mutant was greatly delayed at $25^{\circ}C$ and $36^{\circ}C$, and a large number of multi-septate cells were produced. The mas1 mutant had 2C, 4C and 6C DNA contents, as determined by FACS analysis. In addition, the number of multi-septate cells significantly increased. When cells were cultured in nitrogen starvation medium to increase proliferation, the abnormal phenotypes of mas1 mutant dramatically increased. These phenotypes could be rescued by an overexpression of the $mas1^+$ gene. The mas1 protein localized in the nuclei of S. pombe and human HeLa cells, as evidenced by Mas1-EGFP signals. The abnormal growth pattern and the morphology of mas1 mutant were complemented by a plasmid carrying human CIP29/Hcc-1cDNA. In addition, CIP29 /Hcc-1 transcript level increased in active cell proliferation stages in the developing mouse embryos. These results indicate that the $mas1^+$ ishomologous to the human CIP29/Hcc1 gene and is involved in cytokinesis and cell shape control.

• Journal of Life Science
• /
• v.8 no.3
• /
• pp.279-284
• /
• 1998

The nucleotide sequences of the 5'-untraslated region(5'-UTR) of Hepatitis G virus(HGV) from sera of Korean patients were determines. When compared to the previously reported isolates, the Korean isolates have higher sequence homology with the Japanese isolates indicating the geographic distribution of HGV variants. Interestingly, three discrete regions which are highly conserved among HGV isolates from the same geographical area, thus could be applied to distinguish HGV isolates from the different areas were noticed in the 5'-UTR. Based on the sequences of these group-specific regions, twenty four different HGV isolates could be classified into 5 groups. By using the group-specific regions, inconsistency in HGV typing when based on the different regions of HGV could be solved.

• Research in Plant Disease
• /
• v.17 no.3
• /
• pp.391-395
• /
• 2011

The presence of Pleochaeta shiraiana associated with powdery mildew of Celtis sinensis in Korea is dubious, mainly due to incomplete records of earlier workers. To confirm the occurrence of the powdery mildew and the identity of the causal fungus, morphological characteristics of the anamorph and teleomorph from nine samples deposited in KUS (Herbarium in Korea University) were examined by light microscopy. Identification of the fungus was supported by comparing four ITS sequences from Korean samples with two Japanese data. This is the first confirmed report of the powdery mildew associated with P. shiraiana in Korea.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2007.06a
• /
• pp.477-480
• /
• 2007

얼굴 동작을 효과적으로 인식하는 방법을 제안하고자 한다. 얼굴 동작은 얼굴 표정, 얼굴 자세, 시선, 주름 같은 얼굴 특징이나 얼굴 행동 등으로 표출될 수 있다. 이러한 표출된 정보들은 얼굴 동작이 다양하고 명확하지 않아 연구 진행에 많은 어려움이 있다. 그러므로, 본 논문에서는 얼굴 동작을 묘사하는 FACS를 기반으로 하여 시각적 관찰에 의해 주요한 얼굴 동작을 표현하고, 베이지안 네트워크를 통하여 여러 정보를 분석 융합하여 얼굴 행동을 추론 할 수 있도록 하였다. 베이지안 네트워크의 하향식 추론으로 시각 정보를 선택 할 수 있고, 관측된 현상을 토대로 상향식 추론 하여 얼굴 동작의 신뢰 전파를 통하여 분류 인식한다.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2003.10a
• /
• pp.645-648
• /
• 2003

This paper presents an active vision system for on-line traffic sign recognition. The system is composed of two cameras, one is equipped with a wide-angle lens and the other with a telephoto lends, and a PC with an image processing board. The system first detects candidates for traffic signs in the wide-angle image using color, intensity, and shape information. For each candidate, the telephoto-camera is directed to its predicted position to capture the candidate in a large size in the image. The recognition algorithm is designed by intensively using built in functions of an off-the-shelf mage processing board to realize both easy implementation and fast recognition. The results of on-road experiments show the feasibility of the system.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2003.10a
• /
• pp.236-238
• /
• 2003

Human face detection is often the first step in applications such as video surveillance, human computer interface, fare recognition, and image database management. We have constructed a simple and fast system to detect frontal human faces in complex environment and different illumination. This paper presents a fast segmentation method to combine neighboring pixels with similar hue. The algorithm constructs eye, mouth, and boundary maps for verifying each fare candidate. We test the system on images in complex environment and with confusing objects. The experiment shows a robust detection result with few false detected fates.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2003.10a
• /
• pp.797-800
• /
• 2003

In the current study, the authors propose a method for extracting license plate regions by means of a neural network trained to output the plates center of gravity. The method is shown to be effective. Since the learning pattern presentation positions are defined by random numbers, a different pattern is submitted to the neural network for learning each time, which enables it to form a neural network with high universality of coverage. The article discusses issues of the optimal learning surface for a license plate revered by the learning pattern, the effort of suppression learning of the number and headlight sections, as well as the effect of learning pattern enlargement/reduction and of concentration value conversion. Results of evaluation tests based on pictures of 595 vehicles taken at an underground parking garage demonstrated detection rates of 98.5%.