• 방송공학회논문지
• /
• 제8권2호
• /
• pp.172-180
• /
• 2003

본 논문은 얼굴 모션 캡쳐 애니메이션을 위한 빠르고 정확한 추출 및 추적 알고리즘을 제안한다. 제안한 알고리즘은 추출 및 추적의 두 단계로 구성된다. 먼저 신경회로망 기반의 영역 병합 기법을 이용하여 입력 영상으로부터 다중 마크를 분리한다. 그 다음, 신경회로망 기반의 추적 알고리즘을 사용하여 각각의 프레임에서 추출된 다중 마크들을 추적한다. 실험의 결과는 추출단계에서는 노이즈를 제거하고 처리 시간을 줄일 수 있었다. 또한 낮은 프레임율에서도 성능이 좋은 추적 결과를 얻을 수 있었다.

• 전기전자학회논문지
• /
• 제6권1호
• /
• pp.54-64
• /
• 2002

본 논문에서는 화상회의 영상 데이터에서 화자를 분할하고, 그 움직임을 추적하는 알고리즘을 제안하였다. 실시간 처리가 가능하도록 화자 분할과 움직임 추적 순으로 처리되는 알고리즘으로 단순화하였다. 분할 한계에서는 차분 방법에 의해 구한 움직임 정보와 영상의 밝기 정보를 사용하여 화자를 분할하였다. 분할된 화자로부터 기준 마스크 영상을 생성하였다. 움직임 추적 단계에서는 움직임 추적에 불필요한 블록들은 제외함으로써 빠르게 움직임을 추적할 수 있는 블록정합 알고리즘을 사용하여 추적하였다. 시뮬레이션에서 여러 시험 영상에 제안한 알고리즘을 적용하여 움직이는 화자를 분할하고, 그 움직임를 추적하는 올바른 결과를 얻을 수 있었다.

• 한국항행학회논문지
• /
• 제10권4호
• /
• pp.400-405
• /
• 2006

본 논문에서는 무인항공기 영상에서 움직임 예측을 이용한 이동 물체 추적 알고리즘을 제안하였다. 제안한 알고리즘은 초기모델 생성단계와 이동 물체 추적 단계로 구성되어있으며, 이전 프레임에 비해 이동 거리가 커진 경우에도 안정적으로 추적할 수 있었다. 그리고 무인항공기 카메라의 흔들림에도 효과적으로 추적을 할 수 있었고, 이동 물체의 위치를 정확히 검출하여 추적시간을 단축할 수 있었다. 블록 영상과 참조 영상 간 이동물체의 유사도 판정은 블록 매칭 알고리즘을 사용하였다. 제안한 알고리즘은 실험을 통해서 기존의 전역탐색 알고리즘보다 향상된 결과를 보여주었다.

• 드라이브 ㆍ 컨트롤
• /
• 제16권2호
• /
• pp.22-29
• /
• 2019

This study presented a model predictive tracking control algorithm of autonomous truck based on object state estimation using extended Kalman filter. To design the model, the 1-layer laser scanner was used to estimate position and velocity of the object using extended Kalman filter. Based on these estimations, the desired linear path for object tracking was computed. The lateral and yaw angle errors were computed using the computed linear path and relative positions of the truck. The computed errors were used in the model predictive control algorithm to compute the optimal steering angle for object tracking. The performance evaluation was conducted on Matlab/Simulink environments using planar truck model and actual point data obtained from laser scanner. The evaluation results showed that the tracking control algorithm developed in this study can track the object reasonably based on the model predictive control algorithm based on the estimated states.

• 제어로봇시스템학회:학술대회논문집
• /
• 제어로봇시스템학회 1993년도 한국자동제어학술회의논문집(국내학술편); Seoul National University, Seoul; 20-22 Oct. 1993
• /
• pp.218-223
• /
• 1993

In this paper, we modify a multiple target angle tracking algorithm presented by Sword et al.. The predicted estimates, instead of the existing estimates, of the target angles are updated by the most recent output of the sensor array to improve the tracking performance of the algorithm for crossing targets. Also, the least square solution is modified to avoid abnormally large angle innovations when the target angles are very close. The improved performance of the proposed algorithm is demonstrated by computer simulations.

• Journal of Electrical Engineering and information Science
• /
• 제3권2호
• /
• pp.267-273
• /
• 1998

This paper introduces an adaptive algorithm determining the measurement-track association problem in multi-target tracking(MTT). We model the target and measurement relationships with mean field theory and then define a MAP estimate for the optimal association. Based on this model, we introduce an energy function defined over the measurement space, that incorporates the natural constraints for target tracking. To find the minimizer of the energy function, we derived a new adaptive algorithm by introducing the Lagrange multipliers and local dual theory. Through the experiments, we show that this algorithm is stable and works well in general environments. Also the advantages of the new algorithm over other algorithms are discussed.

• 제어로봇시스템학회:학술대회논문집
• /
• 제어로봇시스템학회 1998년도 제13차 학술회의논문집
• /
• pp.323-328
• /
• 1998

In this paper, a new input estimation algorithm is proposed for target tracking problem. The unknown target maneuver is approximated by a linear combination of independent time functions and the coefficients are estimated by using a weighted least-squares estimation technique. The proposed algorithm is verified by computer simulation of a realistic two-dimensional tracking problem. The proposed algorithm provides significant improvements in estimation performance over the conventional input estimation techniques based on the constant-input assumption.

• 대한전자공학회:학술대회논문집
• /
• 대한전자공학회 2007년도 하계종합학술대회 논문집
• /
• pp.405-406
• /
• 2007

In this paper, we propose feature extraction and tracking algorithm using multi resolution in 2-dimensional wavelet domain. Feature extraction selects feature points using 2-level wavelet transform in interested region. Feature tracking estimates displacement between current frame and next frame based on feature point which is selected feature extraction algorithm. Experimental results show that the proposed algorithm confirmed a better performance than the existing other algorithms.

• 대한전기학회:학술대회논문집
• /
• 대한전기학회 1996년도 하계학술대회 논문집 B
• /
• pp.943-945
• /
• 1996

In this paper, we present an algorithm that allows tracking of a target using measurements obtained from a sensor with limited resolution. The Interacting Multiple Model (IMM) algorithm has been shown to be one of the most cost-effective estimation schemes for hybrid systems. The approach consists of IMM algorithm combined with a coupled version of the Joint Probabilistic Data Association Filter for the target that splits into two targets.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제7권5호
• /
• pp.1166-1179
• /
• 2013

In wireless sensor network (WSN) environments, environmental noises are generated by, for example, small passing animals, crickets chirping or foliage blowing and will interfere target detection if the noises are higher than the sensor threshold value. For accurate tracking by acoustic WSNs, these environmental noises should be filtered out before initiating track. This paper presents the effect of environmental noises on target tracking and proposes a new algorithm for the noise mitigation in acoustic WSNs. We find that our noise mitigation algorithm works well even for targets with sensing range shorter than the sensor separation as well as with longer sensing ranges. It is also found that noise duration at each sensor affects the performance of the algorithm. A detection algorithm is also presented to account for the Doppler effect which is an important consideration for tracking higher-speed ground targets. For tracking, we use the weighted sensor position centroid to represent the target position measurement and use the Kalman filter (KF) for tracking.