• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제4권3호
• /
• pp.358-379
• /
• 2010

Multiple object association is an important capability in visual surveillance system with multiple cameras. In this paper, we introduce locally initiating line-based object association with the parallel projection camera model, which can be applicable to the situation without the common (ground) plane. The parallel projection camera model supports the camera movement (i.e. panning, tilting and zooming) by using the simple table based compensation for non-ideal camera parameters. We propose the threshold distance based homographic line generation algorithm. This takes account of uncertain parameters such as transformation error, height uncertainty of objects and synchronization issue between cameras. Thus, the proposed algorithm associates multiple objects on demand in the surveillance system where the camera movement dynamically changes. We verify the proposed method with actual image frames. Finally, we discuss the strategy to improve the association performance by using the temporal and spatial redundancy.

• 한국멀티미디어학회논문지
• /
• 제20권6호
• /
• pp.893-910
• /
• 2017

In this paper, we propose a simple online multiple object (human) tracking method, LKDeep (Lucas-Kanade feature and Detection based Simple Online Multiple Object Tracker), which can run in fast online enough on CPU core only with acceptable tracking performance for embedded surveillance purpose. The proposed LKDeep is a pragmatic hybrid approach which tracks multiple objects (humans) mainly based on LK features but is compensated by detection on periodic times or on necessity times. Compared to other state-of-the-art multiple object tracking methods based on 'Tracking-By-Detection (TBD)' approach, the proposed LKDeep is faster since it does not have to detect object on every frame and it utilizes simple association rule, but it shows a good object tracking performance. Through experiments in comparison with other multiple object tracking (MOT) methods using the public DPM detector among online state-of-the-art MOT methods reported in MOT challenge [1], it is shown that the proposed simple online MOT method, LKDeep runs faster but with good tracking performance for surveillance purpose. It is further observed through single object tracking (SOT) visual tracker benchmark experiment [2] that LKDeep with an optimized deep learning detector can run in online fast with comparable tracking performance to other state-of-the-art SOT methods.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제4권6호
• /
• pp.1169-1193
• /
• 2010

In this paper, we present and compare two different multiple camera collaboration strategies to reduce false association in finding the correspondence of objects. Collaboration matrices are defined with the required minimum separation for an effective collaboration because homographic lines for objects association are ineffective with the insufficient separation. The first strategy uses the collaboration matrices to select the best pair out of many cameras having the maximum separation to efficiently collaborate on the object association. The association information in selected cameras is propagated to unselected cameras by the global information constructed from the associated targets. While the first strategy requires the long operation time to achieve the high association rate due to the limited view by the best pair, it reduces the computational cost using homographic lines. The second strategy initiates the collaboration process of objects association for all the pairing cases of cameras regardless of the separation. In each collaboration process, only crossed targets by a transformed homographic line from the other collaborating camera generate homographic lines. While the repetitive association processes improve the association performance, the transformation processes of homographic lines increase exponentially. The proposed methods are evaluated with real video sequences and compared in terms of the computational cost and the association performance. The simulation results demonstrate that the proposed methods effectively reduce the false association rate as compared with basic pair-wise collaboration.

• 한국멀티미디어학회논문지
• /
• 제22권2호
• /
• pp.167-177
• /
• 2019

Even though so much progresses have been achieved in Multiple Object Tracking (MOT), most of reported MOT methods are not still satisfactory for commercial embedded products like Pan-Tilt-Zoom (PTZ) camera. In this paper, we propose a real-time multiple pedestrians tracking method for embedded environments. First, we design a new light weight convolutional neural network(CNN)-based pedestrian detector, which is constructed to detect even small size pedestrians, as well. For further saving of processing time, the designed detector is applied for every other frame, and Kalman filter is employed to predict pedestrians' positions in frames where the designed CNN-based detector is not applied. The pose orientation information is incorporated to enhance object association for tracking pedestrians without further computational cost. Through experiments on Nvidia's embedded computing board, Jetson TX2, it is verified that the designed pedestrian detector detects even small size pedestrians fast and well, compared to many state-of-the-art detectors, and that the proposed tracking method can track pedestrians in real-time and show accuracy performance comparably to performances of many state-of-the-art tracking methods, which do not target for operation in embedded systems.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제9권6호
• /
• pp.2217-2229
• /
• 2015

Since the variations of illumination, the irregular changes of human shapes, and the partial occlusions, multiple person tracking is a challenging work in computer vision. In this paper, we propose a graph clustering method based on spatio-temporal information of moving objects for multiple person tracking. First, the part-based model is utilized to localize individual foreground regions in each frame. Then, we heuristically leverage the spatio-temporal constraints to generate a set of reliable tracklets. Finally, the graph shift method is applied to handle tracklet association problem and consequently generate the completed trajectory for individual object. The extensive comparison experiments demonstrate the superiority of the proposed method.

• 한국콘텐츠학회논문지
• /
• 제7권2호
• /
• pp.40-50
• /
• 2007

본 논문에서는 카메라로부터 획득 되어진 비디오 시퀀스로부터 다중 움직임 객체와 배경을 분할하고 시공간 정보에 기반 한 객체 추적 방법을 제안한다. 제안한 방법은 3단계로 구성되어 있다. 먼저 입력 비디오 시퀀스로부터 프레임 사이의 차를 이용한 움직임 영역과 움직임이 존재하지 않는 영역을 구분하여 적응적 경계간을 추출한다. 두 번째는 참조 배경영상과 적응적 경계값을 이용하여 움직임이 존재하는 영역으로부터 개략적 객체 분할을 수행하며, 분할된 이진영상에 형태학적 영역 병합 알고리즘을 적용하여 객체 병합을 수행하였다. 마지막으로 분할된 객체에 시공간 정보를 이용하여 객체에 임의의 ID를 할당하여 추적하였다. 카메라로부터 획득되어진 비디오 시퀀스를 이용한 실험에서 객체들의 분할 및 추적의 효율성과 시스템의 유용성을 확인하였다.

• 한국전자통신학회논문지
• /
• 제13권6호
• /
• pp.1263-1268
• /
• 2018

본 논문에서는 영상 보안 감시를 위한 심층학습 객체 검출과 다중 객체 추적을 위한 확률적 데이터연관 필터를 연계한 영상분석 기법을 제안하고, GPU를 이용하여 구현하는 방안을 제시한다. 제안하는 영상분석 기법은 객체 검출과 추적으로 순차적으로 수행한다. 객체 검출을 위한 심층학습은 ResNet을 이용하고, 다중 객체 추적을 위하여 확률적 데이터 연관 필터를 적용한다. 제안하는 영상분석 기법은 임의의 영역으로 불법으로 침입하는 사람을 검출하거나 특정 공간에 출입하는 사람을 계수하는데 응용할 수 있다. 시뮬레이션을 통하여 약 25fps의 속도로 48채널의 영상을 분석할 수 있음을 보이고, RTSP 프로토콜을 통하여 실시간 영상분석이 가능함을 보인다.

• 대한조선학회논문집
• /
• 제60권4호
• /
• pp.248-255
• /
• 2023

In order to implement the autonomous navigation function, it is essential to track an object within a certain radius of the ship's route. This paper proposes the Multiple Probabilistic Data Association Filter (MPDAF), which can track multiple ships by extending Probabilistic Data Association Filter (PDAF), an existing single object tracking algorithm, using radar data obtained from real marine environments. The proposed MPDAF algorithm was developed to address the problem of tracking multiple objects in a complex environment where there can be significant uncertainty in the number and identification of objects to be tracked. Using real-world radar data provided by the German aerospace center (DLR), it has been verified that the proposed algorithm can track a large number of objects with a small position error.

• ETRI Journal
• /
• 제37권3호
• /
• pp.551-561
• /
• 2015

We propose a novel multiple-object tracking algorithm for real-time intelligent video surveillance. We adopt particle filtering as our tracking framework. Background modeling and subtraction are used to generate a region of interest. A two-step pedestrian detection is employed to reduce the computation time of the algorithm, and an iterative particle repropagation method is proposed to enhance its tracking accuracy. A matching score for greedy data association is proposed to assign the detection results of the two-step pedestrian detector to trackers. Various experimental results demonstrate that the proposed algorithm tracks multiple objects accurately and precisely in real time.

• 한국정보시스템학회지:정보시스템연구
• /
• 제3권
• /
• pp.71-82
• /
• 1994

Knowledge acquisition, maintenance of knowledge base, and validation and verification of knowledge are the addressed bottlenecks of building successful knowledge based systems. Along with the increment of interesting in the knowledge based systems, the organization needs to develop a new one although it has a similar one. This causes several serious problems including knowledge redundancy and maintenance of knowledge base. This paper present three models of the reusable knowledge base which might be the solution to the above problem. Three models are : 1) multiple knowledge bases for a single AI application, 2) multiple knowledge bases for multiple AI applications, 3) a single knowledge base for multiple AI applications. A new approach to build such a reusable knowledge base in a homogeneous environment is presented. Our model combines the essential object-oriented techniques with rules in a consistent manner. Important aspects of applying object-oriented techniques to AI are discussed (inheritance, encapsulation, message passing), and some potential problems in building an AI application (decomposition technique of knowledge, search time, and heterogeneous environment) are pointed out. The models of a reusable knowledge base provide several amenities : 1) reduce the knowledge redundancy, 2) reduce the effort of maintenance of the knowledge base, 3) reuse the resource of the multiple domain knowledge bases, 4) reduce the development time.