• 한국멀티미디어학회논문지
• /
• 제20권7호
• /
• pp.986-993
• /
• 2017

Object Tracking is a technique for tracking moving objects over time in a video image. Using object tracking technique, many research are conducted such a detecting dangerous situation and recognizing the movement of nearby objects in a smart car. However, it still remains a challenging task such as occlusion, deformation, background clutter, illumination variation, etc. In this paper, we propose a novel deep visual object tracking method that can be operated in robust to many challenging task. For the robust visual object tracking, we proposed a Convolutional Neural Network(CNN) which shares weight of the convolutional layers. Input of the CNN is a three; first frame object image, object image in a previous frame, and current search frame containing the object movement. Also we propose a method to consider the motion of the object when determining the current search area to search for the location of the object. Extensive experimental results on a authorized resource database showed that the proposed method outperformed than the conventional methods.

• 한국군사과학기술학회지
• /
• 제14권3호
• /
• pp.405-413
• /
• 2011

Defense Systems Test Center in ADD supports increasingly various missile test requirements such as higher altitude event, multi target operation and low-altitude, high velocity target tracking. In this paper, we have proposed the development of instrumentation radar tracking status simulator based on virtual reality. This simulator can predict the tracking status and risk of failure using several modeling algorithms. It consists of target model, radar model, environment model and several algorithms includes the multipath interference effects. Simulation results show that the predict tracking status and signal are similar to the test results of the live flight test. This simulator predicts and analyze all of the status and critical parameters such as the optimal site location, servo response, optimal flight trajectory, LOS(Line of Sight). This simulator provides the mission plan with a powerful M&S tool to rehearse and analyze instrumentation tracking radar measurement plan for live flight test at DSTC(Defense Systems Test Center).

• 한국항행학회논문지
• /
• 제26권6호
• /
• pp.522-527
• /
• 2022

객체추적은 이전시간에서 추정한 위치와 현재 관측 데이터를 바탕으로 객체의 위치를 연속적으로 추적하는 신호처리 분야이다. 이 논문에서는 3개의 RNN을 서브모듈로 가지는 적응형 스케일조절 신경망을 이용해 입력 데이터의 스케일을 스스로 조절하여 추적할 수 있는 신경망을 제안한다. 객체 추적 성능을 평가하기 위해 객체가 조각별 등가속운동을 하는 1차원 객체 운동 모델에서 제안하는 시스템, 칼만 필터와 최대우도기법의 추적 성능을 비교한다. 그 결과 제안하는 알고리듬의 성능이 평균제곱근오차 기준으로 최대우도기법과 칼만필터보다 다양한 상황에서 전반적으로 우수하며 관측잡음이 커질수록 성능격차가 더 커지는 것을 보인다.

• 한국산학기술학회논문지
• /
• 제10권12호
• /
• pp.3694-3701
• /
• 2009

본 논문에서는 RFID를 이용한 위치 추적 시스템을 제안하고자 한다. 수동형 RFID 태그를 사용자의 신분증에 삽입하여 위치추적 및 출입인증에 사용한다. 리더는 주기적으로 신호를 브로드캐스팅 하며, 리더는 태그의 응답 신호를 받아 사용자의 위치를 파악할 수 있다. 위치추적 방법으로는 신호의 세기에 따라 이동 경로를 파악할 수 있으며, 오래 머문 곳에 대한 위치를 이용하여 관심 분야 파악이 가능하다. 또한 백앤드 서버에 저장된 태그 ID값을 이용하여 보안구역 내 출입인증 시스템으로 활용이 가능하다.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제5권5호
• /
• pp.909-937
• /
• 2011

Energy efficiency is one of the most critical issues in the design of wireless sensor networks. In object-tracking sensor networks, the data storage and query processing should be energy-conserving by decreasing the message complexity. In this paper, a Prediction-based Energy-conserving Approximate StoragE schema (P-EASE) is proposed, which can reduce the query error of EASE by changing its approximate area and adopting predicting model without increasing the cost. In addition, focusing on reducing the unnecessary querying messages, P-EASE enables an optimal query algorithm to taking into consideration to query the proper storage node, i.e., the nearer storage node of the centric storage node and local storage node. The theoretical analysis illuminates the correctness and efficiency of the P-EASE. Simulation experiments are conducted under semi-random walk and random waypoint mobility. Compared to EASE, P-EASE performs better at the query error, message complexity, total energy consumption and hotspot energy consumption. Results have shown that P-EASE is more energy-conserving and has higher location precision than EASE.

• 제어로봇시스템학회논문지
• /
• 제21권1호
• /
• pp.7-13
• /
• 2015

Unmanned air vehicles (UAVs) are getting popular not only as a private usage for the aerial photograph but military usage for the surveillance, reconnaissance and supply missions. For an UAV to successfully achieve these kind of missions, geolocation (localization) must be implied to track an interested target or fly by reference. In this research, we adopted multi-sensor fusion (MSF) algorithm to increase the accuracy of the geolocation and verified the algorithm using two multicopter UAVs. One UAV is equipped with an optical camera, and another UAV is equipped with an optical camera and a laser range finder. Throughout the experiment, we have obtained measurements about a fixed ground target and estimated the target position by a series of coordinate transformations and sequential Kalman filter. The result showed that the MSF has better performance in estimating target location than the case of using single sensor. Moreover, the experimental result implied that multi-sensor geolocation algorithm is able to have further improvements in localization accuracy and feasibility of other complicated applications such as moving target tracking and multiple target tracking.

• 한국산학기술학회논문지
• /
• 제11권1호
• /
• pp.96-100
• /
• 2010

본 논문에서는 적외선 센서와 초음파 센서를 동시에 사용하여 이동하는 표적의 위치를 실시간으로 추적하는 위치추적 시스템을 설계하였다. 위치 추적 시스템은 2 개의 적외선 센서를 이용하여 표적이 감시영역에 침입하는 것을 확인하고, 4 개의 초음파 센서를 이용해서 각 센서에서 표적까지의 거리를 탐지하여 이동하는 표적의 위치좌표 (x,y)를 구하였다. 특히 초음파 센서가 가지고 있는 안테나의 빔폭 특성 때문에 4 개 센서 중 2개 센서에만 표적이 탐지될 경우에도 표적의 위치를 정확하게 결정하는 방법을 제시하였다. 또한 개발된 알고리즘을 설계된 시스템에 장착하여 시험한 결과 감시시스템이 실험실 내에서 실시간 정확하게 구동하는 것을 확인하였다.

• 한국산업정보학회논문지
• /
• 제25권5호
• /
• pp.41-48
• /
• 2020

최근 가상/증강/혼합현실 분야의 이동 플랫폼 시장 수요 커지면서 가상환경을 이용한 다중 체험이 가능한 콘텐츠를 통해 사용자에게 실제 현장과 같은 느낌을 부여하는 체험형 콘텐츠가 주목받고 있다. 본 논문에서는 교육훈련생의 모션 캡쳐를 위한 가상공간 이동플랫폼에서 사용자 위치 추정을 위한 트래커의 추적하는 방법으로, 2차원 영상 평면에 투영된 마커의 좌표를 통한 3차원 교차 공분산의 3D 좌표 추정 방법을 제시한다. 또한, 강체 추적실험을 통해 제안한 알고리즘의 유효성을 검증하여 낮은 해상도의 카메라를 통해서도 3D 좌표 추정이 가능함을 보인다.

• 대한임베디드공학회논문지
• /
• 제17권5호
• /
• pp.281-288
• /
• 2022

According to statistics conducted by the Korea Coast Guard, the number of accidents on small boats under 5 tons is increasing every year. This is because only a small number of people are on board. The previously developed maritime distress and safety systems are not well distributed because passengers must be equipped with additional remote equipment. The purpose of this study is to develop a distress awareness system that recognizes man over-board situations in real time. This study aims to present the part of the passenger tracking system among the small ship's distress awareness situational system that can generate passenger's location information in real time using deep learning based object detection and tracking technologies. The system consisted of the following steps. 1) the passenger location information is generated in the form of Bounding box using its detection model (YOLOv3). 2) Based on the Bounding box data, Deep SORT predicts the Bounding box's position in the next frame of the image with Kalman filter. 3) When the actual Bounding Box is created within the range predicted by Kalman-filter, Deep SORT repeats the process of recognizing it as the same object. 4) If the Bounding box deviates the ship's area or an error occurs in the number of tracking occupant, the system is decided the distress situation and issues an alert. This study is expected to complement the problems of existing technologies and ensure the safety of individuals aboard small boats.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제3권1호
• /
• pp.84-95
• /
• 2009

The ubiquitous smart home is the home of the future, which exploits context information from both the human and the home environment, providing an automatic home service for the human. Human location and motion are the most important contexts in the ubiquitous smart home. In this paper, we present a real-time human tracker that predicts human location and motion for the ubiquitous smart home. The system uses four network cameras for real-time human tracking. This paper explains the architecture of the real-time human tracker, and proposes an algorithm for predicting human location and motion. To detect human location, three kinds of images are used: $IMAGE_1$ - empty room image, $IMAGE_2$ - image of furniture and home appliances, $IMAGE_3$ - image of $IMAGE_2$ and the human. The real-time human tracker decides which specific furniture or home appliance the human is associated with, via analysis of three images, and predicts human motion using a support vector machine (SVM). The performance experiment of the human's location, which uses three images, lasted an average of 0.037 seconds. The SVM feature of human motion recognition is decided from the pixel number by the array line of the moving object. We evaluated each motion 1,000 times. The average accuracy of all types of motion was 86.5%.