• 한국멀티미디어학회논문지
• /
• 제14권10호
• /
• pp.1252-1261
• /
• 2011

본 논문은 내용기반 동영상 분할을 위한 장면의 유사도 패턴 비교 방법을 제안한다. 동영상 장면 전환의 종류는 크게 급진적 전환과 디졸브(dissolve), 페이드인(fade-in), 페이드아웃(fade-out), 와이프 전환(wipe transition)을 포함하는 점진적 전환 형태로 나눌 수 있다. 제안하는 방법은 모든 종류의 장면 전환 검출 문제를 단지 발생 유무의 문제로 간단 정의하고, 장면 전환 종류는 별도로 구분하지 않는다. 장면 전환을 검출하기 위해서는 프레임간의 유사도를 정의해야 한다. 본 논문에서는 장면 내 유사도(within similarity)와 장면 간 유사도(between similarity)를 정의하며 두 유사도의 통계적 패턴 비교를 통하여 최종적으로 장면 전환을 검출하게 된다. 장면 내 유사도와 장면 간 유사도의 비율을 구하는 방법을 통해 플래시라이트나영상 내 물체 움직임에 대한 거짓 양성 검출을 별도의 후처리 과정 없이도 방지할 수 있음을 확인하였다. 프레임의 특징 값으로는 컬러 히스토그램과 프레임 내 평균 화소값을 이용하였다. TREC-2001, TREC-2002 동영상 셋을 포함한 실험 셋에서 성능을 평가한 결과 제안하는 알고리즘의 경우 총 91.84%의 재현율(recall)과 86.43%의 정확도(precision)의 성능을 보임을 확인할 수 있었다.

• 생물정신의학
• /
• 제20권3호
• /
• pp.80-85
• /
• 2013

Objectives Research suggests that content of delusion and hallucination in schizophrenia is influenced by culture and social environment. However, few studies investigated chronological change of delusions and hallucinations within a society. To investigate changes in delusions and hallucinations of schizophrenia according to time, we compared contents of symptoms between inpatients with schizophrenia at two different time frames. Methods All admissions to a psychiatric unit of Hanyang University Guri Hospital with discharge diagnoses of schizophrenia at two different five-year time frames (1996-2000 and 2006-2010) were reviewed. Using a checklist, adapted from the Scale for Assessment of Positive Symptoms, we investigated delusions and hallucinations of 247 patients (104 in the 1990s and 143 in the 2000s). Results Delusions and hallucinations of patients did not differ at two time frames. In women, however, auditory and somato-tactile hallucinations were significantly more frequent in the 1990s (p < 0.05). Conclusions Our findings support the literature emphasizing that content and frequency of hallucination may differ according to cultural environment. We speculate that attitude toward sex and defensiveness toward disclosure of symptoms may have contributed to interval difference.

• 대한전자공학회논문지SP
• /
• 제46권5호
• /
• pp.135-142
• /
• 2009

홀드형 디스플레이 시스템이 개발되면서 화질의 개선을 위해 프레임율 증가 변환 기법이 고려되고 있다. 프레임율 증가 변환 기법은 두 인접 프레임들로부터 하나 이상의 사이 프레임을 보간하여 시간적 해상도를 향상시킨다. 본 논문에서는 회전을 고려하여 양방향 움직임 예측 방법을 적용한 새로운 프레임율 증가 변환 기법을 제안한다. 먼저 각 블록의 정확한 움직임 벡터를 얻기 위해 두 인접 프레임들로부터 양방향 움직임 예측 방법을 수행한다. 초기 움직임 벡터가 정해지면, 움직임 벡터 필드의 연속성과 정확하지 못한 움직임 벡터의 제거를 위해 움직임 벡터 필드에 메디안 필터를 적용한다. 벡터 메디안 필터를 거친 움직임 벡터들은 회전을 고려한 정밀 움직임 보정을 통해 정밀한 움직임 벡터로 갱신된다. 회전요소를 가진 최종 움직임 벡터가 구해지면, 중첩된 블록 이동 보상 방법을 적용하여 중간 프레임을 생성할 수 있다. 실험결과, 제안하는 알고리즘은 이전 방법에 비해 주관적 및 객관적으로 더 좋은 성능을 보였다.

• Steel and Composite Structures
• /
• 제41권5호
• /
• pp.663-679
• /
• 2021

To elucidate the differences in the collapse behavior between a single-story beam-column assembly and multi-story frame, two 1/3-scale two-bay composite frames, including a single-story composite beam-column assembly and a three-story composite sub-frame, were designed and quasi-statically tested. The load-displacement responses, failure modes, and internal force development of the two frames were analyzed and compared in detail. Furthermore, the resistance mechanisms of the two specimens were explored, and the respective contributions of different load-resisting mechanisms to the total resistances were quantitatively separated to gain deeper insights. The experimental tests indicated that Vierendeel action was present in the two-dimensional multi-story frames, which led to an uneven internal force distribution among the three stories. The collapse resistance of TSDWA-3S in the flexural stage was not significantly increased by the structural redundancy provided by the additional story, as compared to that of TSDWA-1S. Although the development of the load response was similar in the two specimens at flexural stage, the collapse mechanisms of the multi-story composite frame were much more complicated than those of the single-story beam-column assembly, and the combined action between stories was critical in determining the internal force redistribution and rebalancing of the remaining structure.

• 국제학술발표논문집
• /
• The 9th International Conference on Construction Engineering and Project Management
• /
• pp.877-885
• /
• 2022

Vision-based activity recognition has been widely attempted at construction sites to estimate productivity and enhance workers' health and safety. Previous studies have focused on extracting an individual worker's postural information from sequential image frames for activity recognition. However, various trades of workers perform different tasks with similar postural patterns, which degrades the performance of activity recognition based on postural information. To this end, this research exploited a concept of human-object interaction, the interaction between a worker and their surrounding objects, considering the fact that trade workers interact with a specific object (e.g., working tools or construction materials) relevant to their trades. This research developed an approach to understand the context from sequential image frames based on four features: posture, object, spatial features, and temporal feature. Both posture and object features were used to analyze the interaction between the worker and the target object, and the other two features were used to detect movements from the entire region of image frames in both temporal and spatial domains. The developed approach used convolutional neural networks (CNN) for feature extractors and activity classifiers and long short-term memory (LSTM) was also used as an activity classifier. The developed approach provided an average accuracy of 85.96% for classifying 12 target construction tasks performed by two trades of workers, which was higher than two benchmark models. This experimental result indicated that integrating a concept of the human-object interaction offers great benefits in activity recognition when various trade workers coexist in a scene.

• Earthquakes and Structures
• /
• 제25권6호
• /
• pp.417-427
• /
• 2023

If the center of mass and center of stiffness or strength of a structure plan do not coincide, the structure is considered asymmetric. During an earthquake, in addition to lateral vibration, the structure experiences torsional vibration as well. Lateraltorsional coupling in asymmetric structures in the plan will increase lateral displacement at the ends of the structure plan and, as a result, uneven deformation demand in seismically resistant frames. The demand for displacement in resistant frames depends on the magnitude of transitional displacement to rotational displacement in the plan and the correlation between these two. With regard to the inability to eliminate the asymmetrical condition due to various reasons, such as architectural issues, this study has attempted to use supplemental viscous dampers to decrease the correlation between lateral and torsional acceleration or displacement in the plan. This results in an almost even demand for lateral deformation and acceleration of seismic resistant frames. On this basis, using the concept of Torsional Balance, adequate distribution of viscous dampers for the decrease of this correlation was determined by transferring the "Empirical Center of Balance" (ECB) to the geometrical center of the structure plan and thus obtaining an equal mean square value of displacement and acceleration of the plan edges. This study analyzed stiff and flexible torsional structures with one-way and two-way mass asymmetry in the Opensees software. By implementing the Particle Swarm Optimization (PSO) algorithm, the optimum formation of dampers for controlling lateral displacement and acceleration is determined. The results indicate that with the appropriate distribution of viscous dampers, not only does the lateral displacement and acceleration of structure edges decrease but the lateral displacement or acceleration of the structure edges also become equal. It is also observed that the optimized center of viscous dampers for control of displacement and acceleration of structure depends on the amount of mass eccentricity, the ratio of uncoupled torsional-to-lateral frequency, and the amount of supplemental damping ratio. Accordingly, distributions of viscous dampers in the structure plan are presented to control the structure's torsion based on the parameters mentioned.

• 한국과학교육학회지
• /
• 제42권5호
• /
• pp.515-524
• /
• 2022

본 연구에서는 예비초등교사의 '과학 창의성'과 '과학 창의성 교육'에 대한 인식의 연관성을 틀 내 및 틀 간 창의성을 중심으로 탐색하였다. 틀 내 창의성은 패러다임 내에서 작동하는 이론적 창의성과 실험적 창의성으로 구분되고, 틀 간 창의성을 패러다임의 변화를 가져오는 이론적 창의성을 의미한다. 자료 수집은 심층 면담으로 이루어졌으며, 분석은 틀 내 및 틀 간 범주를 토대로 수행되었다. 연구결과, 예비초등교사들은 주로 틀 내 창의성의 관점에서 '과학 창의성'을 인식하고 있었다. 또한 이들은 '과학 창의성'을 틀 내 창의성 관점에서 실험적 창의성과 이론적 창의성의 관점에서 다양하게 생각하고 있었다. 반면 예비초등교사들은 '과학 창의성 교육'은 틀 내 실험적 창의성의 관점에서 가능하다고 인식하고 있었다. 이러한 결과를 바탕으로 과학 교육에서 고려할 수 있는 과학 창의성의 특성과 이를 위한 과학교육의 방향에 대해서 논의하고자 한다.

• Structural Engineering and Mechanics
• /
• 제33권2호
• /
• pp.159-177
• /
• 2009

The aim of this paper is to give a rigorous framework for the interpretation of limit analysis results including large displacements. The presentation is oriented towards unidimensional media (beams) but two-dimensional (plates) or three-dimensional media are also concerned. A single-degree-of-freedom system is first considered: it shows the basic phenomena of large displacement limit analysis or second-order limit analysis. The results are compared to those of a continuous system and the differences between both systems are discussed. Theoretical results are obtained using the kinematical approach of limit analysis. An admissible load-displacement plane is then defined, according to the yield design theory. The methodology used is applied to frame structures. The presented results are nevertheless different from those already published in the literature, as the virtual displacement field can be distinguished from the displacement field at collapse. The simplicity of large displacement limit analysis makes it attractive for practical engineering applications. The load-displacement upper bound can be used for instance in the optimal design of steel frames in seismic areas.

• 한국정보통신학회:학술대회논문집
• /
• 한국정보통신학회 2012년도 춘계학술대회
• /
• pp.151-152
• /
• 2012

본 논문은 지역의 의사 결정 트리와 클러스터링을 사용하여 장면 전환 검출 방법을 제시한다. 지역 의사 결정 트리는 클러스터 경계선 검출 장면과 그 인접 프레임의 사이의 차이 값을 시간 유사분포를 비교하기 위해 같은 방식으로 검출하고, 그리고 클러스터 단위로 차이 값의 유사성과 그룹 프레임의 끊어지지 않는 시퀀스를 감지한다.

• 대한전자공학회:학술대회논문집
• /
• 대한전자공학회 2000년도 추계종합학술대회 논문집(4)
• /
• pp.151-154
• /
• 2000

We propose a statically motivated scene change detection algorithm. As the difference between the neighboring frames will generate peaks at scene boundaries, the problem of detecting fast scene changes is equivalent to detecting peaks in a given sequence. In this paper, the peak detection is performed via several statistics, namely the sample means and variances. For eliminating flash lights as well as detecting fast scene changes within a small number of frames, we have opted to use a two-stage process for computing the necessary statistics. The results indicate superiority of necessary statistics. The results indicate superiority of the proposed algorithm over the previously reported algorithm.