• 대한전기학회:학술대회논문집
• /
• 대한전기학회 2003년도 학술회의 논문집 정보 및 제어부문 A
• /
• pp.199-202
• /
• 2003

An effective method is proposed for detecting, acquisition and tracking of a moving object using a disturbance map method in complex image sequences. A significant moving object is detected and tracked within the field of view by computing a modified disturbance map method between an Input image and a temporal average image. This method is very efficient in the serveillance application of digital CCTV and an automatic tracking camera. Experimental results using a real image sequence confirmed that the proposed method can effectively detect and track a significant moving object in complex image sequences.

• 한국게임학회 논문지
• /
• 제12권3호
• /
• pp.87-96
• /
• 2012

본 논문에서는 시퀀스 상에서 확장 칼만필터(Extended Kalman Filter) 기반의 SLAM(Simultaneous Localization And Mapping) 시스템의 안정적인 카메라 추적과 재위치(re-localization) 방법이 제안된다. SLAM으로 얻어진 3차원 특징점에 들로네(Delaunay) 삼각화를 적용하여 기준(reference) 평면을 설정하며, 평면상에 존재하는 특징점의 BRISK(Binary Robust Invariant Scalable Keypoints) 기술자(descriptor)를 생성한다. 기존 확장 칼만필터의 오차가 누적되는 경우를 판단하여 기준 평면의 호모그래피로부터 카메라 정보를 해석한다. 또한 카메라가 급격하게 이동해서 특징점 추적이 실패하면, 저장된 강건한 기술자 정보를 매칭하여 카메라의 위치를 다시 추정한다.

• 한국HCI학회:학술대회논문집
• /
• 한국HCI학회 2008년도 학술대회 1부
• /
• pp.775-778
• /
• 2008

증강현실은 가상현실의 한 분야로 실제 환경에 가장 사물을 합성하여 원래의 환경에 존재하는 사물처럼 보이도록 하는 컴퓨터 그래픽 기법이다. 증강현실은 가상의 공간과 사물만을 대상으로 하는 기존의 가상현실과 달리 현실세계 기반위에 가상의 사물을 합성하여 현실세계 만으로는 얻기 어려운 부가적인 정보를 보강해 제공할 수 있는 특징을 가지고 있다. 실세계 기반위에 가상의 사물의 합성을 구현하는데 있어 중요하게 여겨지는 기반 기술인 레지스트레이션 방법이 있다. 레지스트레이션 방법은 실사영상과 3차원 그래픽 객체의 위치와 방향을 결정하는 방법으로서, 모델기반추적과 Move-Matching방법이 사용되고 있다. 본 논문에서는 모델기반추적방법에 대하여 물체 추적을 통한 물체의 정보와 색상 분포를 이용하여 각 장면간의 물체 추적을 통하여 전환되는 장면에서의 모델을 통해 상대적 좌표계를 생성하는 방법에 대하여 연구하였다.

• International Journal of Fuzzy Logic and Intelligent Systems
• /
• 제11권2호
• /
• pp.118-123
• /
• 2011

Video based object tracking normally deals with non-stationary image streams that change over time. Robust and real time moving object tracking is considered to be a problematic issue in computer vision. Multiple object tracking has many practical applications in scene analysis for automated surveillance. In this paper, we introduce a specified object tracking based particle filter used in an environment of multiple moving objects. A differential image region based tracking method for the detection of multiple moving objects is used. In order to ensure accurate object detection in an unconstrained environment, a background image update method is used. In addition, there exist problems in tracking a particular object through a video sequence, which cannot rely only on image processing techniques. For this, a probabilistic framework is used. Our proposed particle filter has been proved to be robust in dealing with nonlinear and non-Gaussian problems. The particle filter provides a robust object tracking framework under ambiguity conditions and greatly improves the estimation accuracy for complicated tracking problems.

• 대한전자공학회:학술대회논문집
• /
• 대한전자공학회 2002년도 ITC-CSCC -1
• /
• pp.200-203
• /
• 2002

We suggest a novel approach to the tracking of multiple moving objects in image sequence. The tracking of multiple moving objects include some complex problems(crossing (occluding), entering, disappearing, joining, and dividing) for objects identifying. Our method can settle these problems by optimization of a penalty function and movement prediction. It is executable in .eat time processing (more than 30 ㎐) because it is computed by only location data.

• 대한전자공학회:학술대회논문집
• /
• 대한전자공학회 2002년도 하계종합학술대회 논문집(4)
• /
• pp.137-140
• /
• 2002

This paper presents a new variational framework for detecting and tracking moving objects in image sequence. Motion detection is performed using Level Set Model. The original frame is used to provide th moving object boundaries Then, the detection and the tracking problem are addressed in a common framework that employs a inward-outward curve evolution function. This function is minimized using a gradient decent method.

• 스마트미디어저널
• /
• 제5권1호
• /
• pp.78-87
• /
• 2016

The major problems of recent object tracking methods are related to the inefficient detection of moving objects due to occlusions, noisy background and inconsistent body motion. This paper presents a robust method for the detection and tracking of a moving in infrared animal videos. The tracking system is based on adaptive optical flow generation, Gaussian mixture and Kalman filtering. The adaptive Gaussian model of optical flow (GMOF) is used to extract foreground and noises are removed based on the object motion. Kalman filter enables the prediction of the object position in the presence of partial occlusions, and changes the size of the animal detected automatically along the image sequence. The presented method is evaluated in various environments of unstable background because of winds, and illuminations changes. The results show that our approach is more robust to background noises and performs better than previous methods.

• 대한공간정보학회지
• /
• 제18권2호
• /
• pp.93-97
• /
• 2010

본 연구에서는 카메라 자세 정보가 없는 무인헬기에 탑재된 카메라로부터 취득된 연속영상을 등록하기 위한 개량형 KLT기법을 제시하였으며 그 절차는 다음과 같이 구성된다. 초기 특징점은 연속영상에서 모서리점을 검출하고 동적프로그래밍에 의한 특성곡선매칭에 의해 특징점을 추적하였다. 추적된 특징점 중 오류점은 RANSAC추정법에 의해 제거되며 호모그래피이론에 의해 나머지점은 정확한 정합점으로 분류되었다. 영상등록에 의한 편위보정영상모자이크생성은 쌍일차보간법에 의해 생성하였으며, 결과분석을 통해 제시된 방법이 흔들림이 있는 연속영상을 등록하는데 적합한 방법임을 제시하였다.

• 한국멀티미디어학회논문지
• /
• 제23권10호
• /
• pp.1236-1249
• /
• 2020

In a most recent object tracking research work, applying Convolutional Neural Network and Recurrent Neural Network-based strategies become relevant for resolving the noticeable challenges in it, like, occlusion, motion, object, and camera viewpoint variations, changing several targets, lighting variations. In this paper, the LSTM Network-based Tracking association method has proposed where the technique capable of real-time multi-object tracking by creating one of the useful LSTM networks that associated with tracking, which supports the long term tracking along with solving challenges. The LSTM network is a different neural network defined in Keras as a sequence of layers, where the Sequential classes would be a container for these layers. This purposing network structure builds with the integration of tracking association on Keras neural-network library. The tracking process has been associated with the LSTM Network feature learning output and obtained outstanding real-time detection and tracking performance. In this work, the main focus was learning trackable objects locations, appearance, and motion details, then predicting the feature location of objects on boxes according to their initial position. The performance of the joint object tracking system has shown that the LSTM network is more powerful and capable of working on a real-time multi-object tracking process.

• Journal of Information Processing Systems
• /
• 제7권3호
• /
• pp.519-530
• /
• 2011

A Snake is an active contour for representing object contours. Traditional snake algorithms are often used to represent the contour of a single object. However, if there is more than one object in the image, the snake model must be adaptive to determine the corresponding contour of each object. Also, the previous initialized snake contours risk getting the wrong results when tracking multiple objects in successive frames due to the weak topology changes. To overcome this problem, in this paper, we present a new snake method for efficiently tracking contours of multiple objects. Our proposed algorithm can provide a straightforward approach for snake contour rapid splitting and connection, which usually cannot be gracefully handled by traditional snakes. Experimental results of various test sequence images with multiple objects have shown good performance, which proves that the proposed method is both effective and accurate.