• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제6권9호
• /
• pp.2170-2190
• /
• 2012

There are nowadays strong demands for intelligent surveillance systems, which can infer or understand more complex behavior. The application of crowd density estimation methods could lead to a better understanding of crowd behavior, improved design of the built environment, and increased pedestrian safety. In this paper, we propose a new crowd density estimation method, which aims at estimating not only a moving crowd, but also a stationary crowd, using images captured from surveillance cameras situated in various public locations. The crowd density of the moving people is measured, based on the moving area during a specified time period. The moving area is defined as the area where the magnitude of the accumulated optical flow exceeds a predefined threshold. In contrast, the stationary crowd density is estimated from the coarseness of textures, under the assumption that each person can be regarded as a textural unit. A multilayer neural network is designed, to classify crowd density levels into 5 classes. Finally, the proposed method is experimented with PETS 2009 and the platform of Gangnam subway station image sequences.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제12권8호
• /
• pp.3769-3789
• /
• 2018

Social pedestrian groups are the basic elements that constitute a crowd; therefore, detection of such groups is scientifically important for modeling social behavior, as well as practically useful for crowd video understanding. A social group refers to a cluster of members who tend to keep similar motion state for a sustained period of time. One of the main challenges of social group detection arises from the complex dynamic variations of crowd patterns. Therefore, most works model dynamic groups to analysis the crowd behavior, ignoring the existence of stationary groups in crowd scene. However, in this paper, we propose a novel unified framework for detecting social pedestrian groups in crowd videos, including dynamic and stationary pedestrian groups, based on spatiotemporal-oriented energy measurements. Dynamic pedestrian groups are hierarchically clustered based on energy flow similarities and trajectory motion correlations between the atomic groups extracted from principal spatiotemporal-oriented energies. Furthermore, the probability distribution of static spatiotemporal-oriented energies is modeled to detect stationary pedestrian groups. Extensive experiments on challenging datasets demonstrate that our method can achieve superior results for social pedestrian group detection and crowd video classification.

• 한국인터넷방송통신학회논문지
• /
• 제12권2호
• /
• pp.103-110
• /
• 2012

수동적인 보안감시 시스템의 문제점이 계속적으로 제기되면서 실시간으로 공공장소에서의 군중에 대한 관리 및 감독을 지원하는 자동화되고 지능적인 군중 밀도 측정에 대한 필요성이 증대되고 있다. 이에 따라, 군중의 밀도를 측정하기 위한 많은 연구가 시도되었으나 실시간 혼잡도 정보 취득이 어렵고, 조명변화 등에 취약한 한계가 드러났다. 본 논문에서는 이러한 문제점을 해결하기 위해 군중 특징 정보로써 옵티컬 플로우를 검출하고 또한 Sobel 외곽선 추출 알고리즘에 의해 외곽선을 추출하여 각 특징을 입력으로 학습된 다층 신경망을 통해 실시간으로 실외 공공장소에서의 군중 밀도를 측정하였다.

• 한국산학기술학회논문지
• /
• 제21권10호
• /
• pp.415-420
• /
• 2020

군중 유동은 대도시의 철도 환승역, 터미널, 복합 다중 건물, 경기장 등에서 흔히 볼 수 있으며, 이러한 시설물에서의 이용객들의 원활한 흐름 뿐만아니라 안전 확보측면에서도 중요한 요소이다. 본 연구에서는 새로운 군중 유동 해석법을 개발하여 철도 환승 연결로 모델에 대하여 적용하였다. 해석법에서는 출구의 포텐셜 값을 가장 작은 값으로 입력하고, 주변 격자들의 포텐셜 값은 점진적으로 증가시켜서 전체적인 포텐셜 지도를 구성한다. 포텐셜 값이 큰 격자에서 작은 격자로 이동하는 방향 벡터를 구하여 이를 따르는 유적선을 구한다. 이 유적선이 여객 유동의 기본 경로가 된다. 해석 대상의 모든 모델에서, 보행자들은 처음 예측된 최단 거리 경로로 이동하지 않고, 시시각각의 상황에 따라 변경된 대체 경로를 이용하여 이동하였다. 양 방향의 보행자가 서로 마주치는 병목 구역에서도 진입 시차를 두어 분산시키면 보행이 훨씬 더 원활하게 되었다. 이상의 해석 결과로부터, 철도역의 하드웨어적 개량 공사를 하지 않고, 여객 유동 분석과 같은 소프트웨어적 해석으로도 혼잡 완화 방안을 찾을 수 있음을 보여준다.

• 한국안전학회지
• /
• 제33권6호
• /
• pp.123-127
• /
• 2018

Recent catastrophic accidents at the underground subway stations in South Korea have proven that the subway evacuation is an important safety concern. Previous studies have used commercial programs for safety assessment or have been focused on development of computing algorithms rather than the basic analysis data which form the foundation of studies. In this study, we designed a new movement recording apparatus which measured and analyzed crowd movements including but not limited to moving velocity, specific flow rate and crowd density. Moreover, We propose new effective analysis method for evacuation studies with this apparatus.

• 한국통신학회논문지
• /
• 제39C권8호
• /
• pp.731-737
• /
• 2014

본 논문에서는 감시 카메라 환경에서 발생할 수 있는 군중의 비정상 행동 검출 방법을 제안한다. 군중들의 비정상 행동을 산발적으로 퍼지면서 뛰는 행동, 한쪽 방향으로 갑자기 뛰는 행동 두 가지로 정의하였다. 이를 검출하기 위하여 영상에서 움직임 벡터를 추출하여 군중의 비정상 행동 검출에 적합한 서술자 MHOF(Multi-scale Histogram of Optical Flow)와 DCHOF(Directional Change Histogram of Optical Flow)제안하였으며, 이를 이진 분류기인 SVM(Support Vector Machine)을 이용하여 검출하였다. 제안한 방법은 공개 데이터셋인 UMN 데이터와 PETS 2009 데이터를 이용하여 성능을 평가하였고 다른 방법론과의 비교를 통해 제안하는 알고리즘의 우수성을 입증하였다.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제9권8호
• /
• pp.2948-2963
• /
• 2015

In the field of computer vision, visual surveillance systems have recently become an important research topic. Growth in this area is being driven by both the increase in the availability of inexpensive computing devices and image sensors as well as the general inefficiency of manual surveillance and monitoring. In particular, the ultimate goal for many visual surveillance systems is to provide automatic activity recognition for events at a given site. A higher level of understanding of these activities requires certain lower-level computer vision tasks to be performed. So in this paper, we propose an intelligent activity recognition model that uses a structure learning method and a classification method. The structure learning method is provided as a K2-learning algorithm that generates Bayesian networks of causal relationships between sensors for a given activity. The statistical characteristics of the sensor values and the topological characteristics of the generated graphs are learned for each activity, and then a neural network is designed to classify the current activity according to the features extracted from the multiple sensor values that have been collected. Finally, the proposed method is implemented and tested by using PETS2013 benchmark data.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제6권12호
• /
• pp.3315-3337
• /
• 2012

Crowded scenes analysis is a challenging topic in computer vision field. How to detect diverse motion patterns in crowded scenarios from videos is the critical yet hard part of this problem. In this paper, we propose a novel approach to mining motion patterns by utilizing motion information during both long-term period and short interval simultaneously. To capture long-term motions effectively, we introduce Motion History Image (MHI) representation to access to the global perspective about the crowd motion. The combination of MHI and optical flow, which is used to get instant motion information, gives rise to discriminative spatial-temporal motion features. Benefitting from the robustness and efficiency of the novel motion representation, the following motion pattern mining is implemented in a completely unsupervised way. The motion vectors are clustered hierarchically through automatic hierarchical clustering algorithm building on the basis of graphic model. This method overcomes the instability of optical flow in dealing with time continuity in crowded scenes. The results of clustering reveal the situations of motion pattern distribution in current crowded videos. To validate the performance of the proposed approach, we conduct experimental evaluations on some challenging videos including vehicles and pedestrians. The reliable detection results demonstrate the effectiveness of our approach.

• 한국화재소방학회논문지
• /
• 제17권4호
• /
• pp.1-6
• /
• 2003

본 논문에서는 우리나라 초등학생을 대상으로 피난상황 시 교내에서의 여러 가지 이동속도를 측정하였으며 이들의 분석방법에 대하여 논하였다. 여기서 이동속도라 함은 교실 문에서의 유출속도, 복도에서의 개인별 이동속도, 복도에서의 밀도별 이동속도(Crowd movement. Flow veloctiy) 및 개인별 계단에서의 이동 속도 등을 포함하게 된다. 이를 위하여 대전의 한 초등학교를 선택하고 각 학년 남여 각각 15명씩 총 180명을 추출하여 시험에 임하였다. 이를 통하여 초등학교 아동들의 이동속도에 대한 기초 자료를 얻을 수 있었으며 이들 자료를 피난 시뮬레이션을 행할 때 초등학생에 대한 지표로 사용하고자 한다.

• 한국기계가공학회지
• /
• 제15권4호
• /
• pp.30-39
• /
• 2016

There have been many cases of deaths from crushing caused by dense crowds. Numerous studies about pedestrian flow have performed various simulations, but the experimental data to prove the simulations are still not enough. In this paper, the evacuation of pedestrians for proving pedestrian flow simulation is observed. Due to the possibility of real casualties, it is difficult to experiment with humans directly. Therefore, ten C57BL/6NCrSIc mice have been used. It is assumed that C57BL/6NCrSIc mice act like humans in panic situations. Electrical Stimulus Experiments on mice are conducted for exits with various angles. ICY software is applied in this paper. As a result, the mice escape fast at a proper angle of 45 to 60 degrees.