• 한국위성정보통신학회논문지
• /
• 제11권4호
• /
• pp.81-88
• /
• 2016

본 논문은 투과성이 좋고 실내 환경에 강건하며, 수십 센티미터(cm) 급의 고정밀 측위가 가능하다는 점에서 주목 받고 있는 IR-UWB(Impulse-Radio Ultra Wideband) 레이더 시스템에서 신호의 특성을 이용한 적응형 다중 목표물 추정 알고리즘을 제안한다. 목표물에 의해 반사되는 신호는 Peak를 갖는 다는 특성으로 다중의 Peak를 추정하는 알고리즘을 제안하였다. 이러한 알고리즘의 성능을 확인하기 위해서 레이더 앞에 다중 목표물을 두고 기존의 기법과 다중 목표물 추정 알고리즘을 비교하였다. 하나의 송신 안테나와 수신안테나로 목표물들의 위치를 실시간으로 추정한다. 기존의 최고 신호 도출 방식에 비해 추정할 수 있는 수가 늘어나고 다중으로 목표물 도출이 가능하다. 기존의 기법은 하나의 목표물만 추정하다보니 평균 제곱 오차가 1이 나오는 반면 다중 목표물 추정 알고리즘은 약 0.05의 결과가 도출된다. 본 논문에서 제시한 기법은 하나의 IR-UWB 모듈 환경에서 다중의 목표물을 추정 및 응용에 적용할 수 있을 것이라 기대된다.

• 한국군사과학기술학회지
• /
• 제14권3호
• /
• pp.517-523
• /
• 2011

The estimation filter in radar systems must track targets' position within low tracking error. In the Multi-Function Radar(MFR), ${\alpha}-{\beta}$ filter and Kalman filter are widely used to track single or multiple targets. However, due to target maneuvering, these filters may not reduce tracking error, therefore, may lost target tracks. In this paper, a target tracking filter based on particle filtering algorithm is proposed for the MFR. The advantage of this method is that it can track targets within low tracking error while targets maneuver and reduce impoverishment of particles by the proposed resampling method. From the simulation results, the improved tracking performance is obtained by the proposed filtering algorithm.

• 제어로봇시스템학회논문지
• /
• 제17권8호
• /
• pp.783-789
• /
• 2011

Target tracking performance improvement using multi-sensor data fusion is a challenging work. However, biases in the measurements should be removed before various data fusion techniques are applied. In this paper, a bias removing algorithm using measurement data from multi-radar tracking systems is proposed and evaluated by computer simulation. To predict bias estimation performance in various geometric relations between the radar systems and target, a system observability index is proposed and tested via computer simulation results. It is also studied that target tracking which utilizes multi-sensor data fusion with bias-removed measurements results in better performance.

• 한국멀티미디어학회논문지
• /
• 제24권5호
• /
• pp.659-666
• /
• 2021

With the continuous development of deep learning, human behavior recognition algorithms have achieved good results. However, in a multi-person recognition environment, the complex behavior environment poses a great challenge to the efficiency of recognition. To this end, this paper proposes a multi-person pose estimation model. First of all, the human detectors in the top-down framework mostly use the two-stage target detection model, which runs slow down. The single-stage YOLOv3 target detection model is used to effectively improve the running speed and the generalization of the model. Depth separable convolution, which further improves the speed of target detection and improves the model's ability to extract target proposed regions; Secondly, based on the feature pyramid network combined with context semantic information in the pose estimation model, the OHEM algorithm is used to solve difficult key point detection problems, and the accuracy of multi-person pose estimation is improved; Finally, the Euclidean distance is used to calculate the spatial distance between key points, to determine the similarity of postures in the frame, and to eliminate redundant postures.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제10권10호
• /
• pp.5095-5111
• /
• 2016

The Probability Hypothesis Density (PHD) filter is a suboptimal approximation and tractable alternative to the multi-target Bayesian filter based on random finite sets. However, the PHD filter fails to track newborn targets when the target birth intensity is unknown prior to tracking. In this paper, a dual detection-guided newborn target intensity PHD algorithm is developed to solve the problem, where two schemes, namely, a newborn target intensity estimation scheme and improved measurement-driven scheme, are proposed. First, the newborn target intensity estimation scheme, consisting of the Dirichlet distribution with the negative exponent parameter and target velocity feature, is used to recursively estimate the target birth intensity. Then, an improved measurement-driven scheme is introduced to reduce the errors of the estimated number of targets and computational load. Simulation results demonstrate that the proposed algorithm can achieve good performance in terms of target states, target number and computational load when the newborn target intensity is not predefined in multi-target tracking systems.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제16권7호
• /
• pp.2328-2344
• /
• 2022

2D human pose estimation still faces difficulty in low-resolution images. Most existing top-down approaches scale up the target human bonding box images to the large size and insert the scaled image into the network. Due to up-sampling, artifacts occur in the low-resolution target images, and the degraded images adversely affect the accurate estimation of the joint positions. To address this issue, we propose a multi-resolution input feature fusion network for human pose estimation. Specifically, the bounding box image of the target human is rescaled to multiple input images of various sizes, and the features extracted from the multiple images are fused in the network. Moreover, we introduce a guiding channel which induces the multi-resolution input features to alternatively affect the network according to the resolution of the target image. We conduct experiments on MS COCO dataset which is a representative dataset for 2D human pose estimation, where our method achieves superior performance compared to the strong baseline HRNet and the previous state-of-the-art methods.

• 한국컴퓨터정보학회논문지
• /
• 제22권7호
• /
• pp.47-54
• /
• 2017

Recently, correlation filter based trackers have shown excellent tracking performance and computational efficiency. In order to enhance tracking performance in the correlation filter based tracker, search area which is image patch for finding target must include target. In this paper, two methods to discriminatively represent target in the search area are proposed. Firstly, search area location is estimated using pyramidal Lucas-Kanade algorithm. By estimating search area location before filtering, fast motion target can be included in the search area. Secondly, we investigate multi-channel Local Edge Pattern(LEP) which is insensitive to illumination and noise variation. Qualitative and quantitative experiments are performed with eight dataset, which includes ground truth. In comparison with method without search area estimation, our approach retain tracking for the fast motion target. Additionally, the proposed multi-channel LEP improves discriminative performance compare to existing features.

• 융합신호처리학회논문지
• /
• 제14권1호
• /
• pp.33-38
• /
• 2013

본 논문은 적외선 연속 영상에서 배경 추정 필터와 칼만 필터, 평균 이동 알고리즘을 사용하여 다중 소형 표적들의 소멸과 생성시에도 표적들의 위치를 추적하는 시스템을 제안한다. 배경 추정 영상파 원 영상과의 차 영상을 사용해서 정지 영상에서의 표적 후 정보를 구하고, 칼만 필터와 후보 표적의 분류를 이용하여 다중 표적을 추적 한다. 마지막으로 평균 이동 알고리즘을 사용하여 표적들의 세부 위치를 조정한다. 실험을 통하여 배경 추정 필터들의 성능을 비교 분석하였고, 제안하는 알고리즘이 기존의 추적 시스템과 비교하여 안정적으로 추적이 됨을 확인하였다.

• 한국인터넷방송통신학회논문지
• /
• 제13권1호
• /
• pp.89-93
• /
• 2013

본 연구는 공간상에서 수신기에 입사하는 목표물 도래방향을 추정하는 것이다. 본 연구는 다중 입력 다중 출력 배열 안테나 시스템에서 상관성 간섭신호를 제거하기 위해서 제한 행렬을 이용하여 공분산 벡터를 제시하였고 비용함수와 최소 분산방법을 이용하여 목표물 도래방향을 추정하는 알고리즘을 제안하였다. 모의실험을 통하여 본 연구에서 제안된 알고리즘 성능을 기존 SPT_LCMV알고리즘과 비교분석하였다. 목표물 도래방향 추정에서 본 연구에서 제안한 방법이 기존 SPT_LCMV알고리즘보다 우수함을 나타내었다.

• 로봇학회논문지
• /
• 제12권1호
• /
• pp.33-41
• /
• 2017

We present a region-based approach for accurate pose estimation of small mechanical components. Our algorithm consists of two key phases: Multi-view object co-segmentation and pose estimation. In the first phase, we explain an automatic method to extract binary masks of a target object captured from multiple viewpoints. For initialization, we assume the target object is bounded by the convex volume of interest defined by a few user inputs. The co-segmented target object shares the same geometric representation in space, and has distinctive color models from those of the backgrounds. In the second phase, we retrieve a 3D model instance with correct upright orientation, and estimate a relative pose of the object observed from images. Our energy function, combining region and boundary terms for the proposed measures, maximizes the overlapping regions and boundaries between the multi-view co-segmentations and projected masks of the reference model. Based on high-quality co-segmentations consistent across all different viewpoints, our final results are accurate model indices and pose parameters of the extracted object. We demonstrate the effectiveness of the proposed method using various examples.