Earticle

이동 객체에 대한 추적 기술은 최근 중요성이 강조되고 있는 동영상 이해에서 가장 핵심적인 기술의 하나라 할 수있다. 하지만, 동영상이 가지는 조명의 불안정, 객체의 크기나 형태 변화, 카메라 움직임, 그리고 중첩 등으로 인해동영상 내의 이동 객체 추적은 많은 어려움을 가지고 있다. 객체 추적의 가장 대표적인 종래의 방법인 칼만 필터와파티클 필터의 문제점을 개선하는 방법으로 스웜 기반의 방법이 제안되어 있으나 동적으로 변화하는 이동 객체의 특징을 반영하는 개선된 알고리즘이 요구된다. 본 논문에서는 이러한 특징을 반영하여 파티클 스웜 최적화 방법에서사용되는 파라미터 중 가중치 값을 동적으로 변화하는 적응적 파라미터 제어 방법을 제안한다. 각 파티클을 특성에따라 3가지 종류로 구분하고 각각 서로 다른 가중치 값을 부여하는 방식으로 객체 추적의 성능을 개선할 수 있다. 제안된 알고리즘의 적용 결과 중첩 또는 예측하기 어려운 움직임 등과 같은 객체의 비선형적인 움직임이 있는 동영상에 대해 기존 파티클 스웜 방식에 비해 현저한 성능 개선을 보이는 것을 확인할 수 있었다.

Moving object tracking techniques can be considered as one of the most essential technique in the video understanding of which the importance is much more emphasized recently. However, irregularity of light condition in the video, variations in shape and size of object, camera motion, and occlusion make it difficult to tracking moving object in the video. Swarm based methods are developed to improve the performance of Kalman filter and particle filter which are known as the most representative conventional methods, but these methods also need to consider dynamic property of moving object. This paper proposes adaptive parameter control method which can dynamically change weight value among parameters in particle swarm optimization. The proposed method classifies each particle to 3 groups, and assigns different weight values to improve object tracking performance. Experimental results show that our scheme shows considerable improvement of performance in tracking objects which have nonlinear movements such as occlusion or unexpected movement.

Firstly, through the principle analysis and simulation experiment, the maneuvering target tracking algorithm of curve model interacting multiple model tracking algorithm was given. Because the algorithm is simple structure and high cost efficiency, it becomes generally applicable algorithm for the curve tracking model. But, the target mobility is very high in practice, Single target tracking model is no longer applicable curve tracking model. To improve the accuracy of tracking, the adaptive grid interacting multiple model (AGIMM) algorithm was given. The algorithm has two fatal weaknesses in the practical application. First, in maneuvering target tracking process, when the model changes and gradual change, the tracking precision is not high; Second, because the changing model structure is very large model sets, the algorithm is complexity and system processing speed is very slow, which cannot be widely used. To improve the accuracy and its scope of application of the algorithm, The paper proposed the adaptive Kalman filter adaptive interacting multiple model algorithm (AKFAIMM).The algorithm introduced the parameter in the adaptive Kalman filter, and adjusted parameter in maneuvering target tracking, the parameter was adjusted continuously in the curve motion model, it could greatly improve the tracking precision and the application of the model. Second, to improve the algorithm complexity. The paper improved that the angular velocity estimation method replaced centripetal acceleration estimation method on turning curve. The estimation method reduced the number of model set and reduced greatly of computation. At the same time, according to the algorithm in the model changes, the centripetal acceleration could be continuously adjusted and improved the adaptability of the model. The algorithm improved maneuvering target tracking algorithm accuracy. The effectiveness of algorithm was proved the validity by simulation.

본 논문은 계산량이 O( N)인 새로운 형태의 QRD-LS 적응 알고리즘을 제안한다. 제안한 알고리즘의 주요 사항은 입력벡터의 단위벡터 계산이 Given Rotation 과정에서 이루어짐에 근거하고 있다. 알고리즘의 성능 평가는 FIR 시스템 식별 문제를 컴퓨터 시뮬레이션을 통하여 수행하였다. 이 알고리즘은 시뮬레이션의 결과 좋은 성능을 나타내었다. 그리고 이론적으로 평균 측면에서 알고리즘이 최적 계수 벡터에 수렴함을 보였다.

This paper proposes a new QRD-LS adaptive algorithm with computational complexity of O( N). The main idea of proposed algorithm(D-QR-RLS) is based on the fact that the computation for the unit vector of is made from the process during Givens Rotation. The performance of the algorithm is evaluated through computer simulation of FIR system identification problem. As verified by simulation results, this algorithm exhibits a good performance. And, we can see the proposed algorithm converges to optimal coefficient vector theoretically.

This paper deals with RLS algorithm using Newton-Raphson method based adaptive forgetting factor for a passive telemetry RF sensor system in order to estimate the time variant parameter to be included in RF sensor model.

This paper deals with RLS algorithm using Newton-Raphson method based adaptive forgetting factor for a passive telemetry RF sensor system in order to estimate the time variant parameter to be included in RF sensor model. For this estimation with RLS algorithm, phasor typed RF sensor system modelled with inductive coupling principle is used. Instead of applying constant forgetting factor to estimate time variant parameter, the adaptive forgetting factor based on Newton-Raphson method is applied to RLS algorithm without constant forgetting factor to be determined intuitively. Finally, we provide numerical examples to evaluate the feasibility and generality of the proposed method in this paper.

This paper proposes a new algorithm to estimate time-varying parameters by combining KFSM(Kalman Filter with Shift Matrix) with neural network compansator. While the time varying parameters are estimated from KFSM, the error coverence of system, R(k) are compansated by neural network concurrently. The casestudy using computer simulation proves the usefullness and advantages of the proposed algorithm in this paper.