• 한국정보통신학회논문지
• /
• 제24권4호
• /
• pp.494-500
• /
• 2020

본 논문에서는 초광대역 (Ultra-wideband, UWB) 시스템에서 실내 위치 측위를 위한 새로운 거리 추정 기법을 제안한다. 제안하는 기법은 딥러닝 기법 중 하나인 순환 신경망 (RNN)을 기반으로 한다. 순환신경망은 시계열 신호를 처리하는데 유용한데 UWB 신호 역시 시계열 데이터로 볼 수 있기 때문에 순환신경망을 사용한다. 구체적으로, UWB 신호가 IEEE 802.15.4a 실내 채널모델을 통과하고 수신된 신호에서 순환신경망 회귀를 통해 송신기와 수신기 사이의 거리를 추정하도록 학습한다. 이렇게 학습된 순환신경망 모델의 성능은 새로운 수신신호를 이용하여 검증하며 기존의 임계값 기반의 거리 추정 기법과도 비교한다. 성능지표로는 제곱근 평균추정에러 (root mean square error, RMSE)를 사용한다. 컴퓨터 모의실험 결과에 따르면 제안하는 거리 추정 기법은 수신신호의 신호 대 잡음비 (signal to noise ratio, SNR) 및 송수신기 사이의 거리와 상관없이 기존 기법보다 항상 월등히 우수한 성능을 보인다.

• 한국정밀공학회지
• /
• 제30권4호
• /
• pp.383-389
• /
• 2013

Demand for high-rising building has arisen. However, its maintenance is usually executed by labour. It could have a severe problem. We proposed a gondola robot to solve it. In this paper, we designed a height estimation sensor for this gondola. It is consist of pan-tilt unit, ARS sensor, and laser sensor. The pan-tilt unit keeps the laser sensor to indicate the gravity direction by referencing the ARS. The laser sensor's range is vertical distance from gondola to ground. However, if there is an obstacle under the gondola, the distance includes its height. To filter it out, we apply a Kalman filter for the height estimation. If the estimated height is changed extremely, the filter decides that there is an obstacle. Then, it remembers the height of obstacle. Other extreme changes of height estimations are reflected. The experimental results using the proposed sensor system show detail flow of the height estimation.

• 한국정보통신학회논문지
• /
• 제14권2호
• /
• pp.469-478
• /
• 2010

본 논문은 3차원 발 자세를 추정하기 위한 효과적 형상 기술자를 제안하였다. 처리 시간을 단축시키기 위하여 특수 제작된 3차원 발 모형을 2차원 투영하여 발 형상 데이터베이스를 구축하고, 3차원 자세 요약정보를 메타 정보로 추가한 2.5차원 영상 데이터베이스를 구성하였다. 그리고 특징 공간 크기가 작고 다른 형상 기술자에 비하여 자세 추정 성능이 뛰어난 수정된 Centroid Contour Distance를 제안하였다. 제안된 기술자의 성능을 분석하기 위하여, 검색 정확도와 시공간 복잡도를 계산하고 기존의 방식들과 비교하였다. 실험 결과를 통하여 제안된 기술자는 특징 추출 시간과 자세 추정 정확도면에서 기존의 방식들보다 효과적인 것으로 나타났다.

• 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.

• 대한조선학회논문집
• /
• 제58권6호
• /
• pp.424-430
• /
• 2021

Fire localization is a key mission that must be preceded for an autonomous fire suppression system. Although studies using a variety of sensors for the localization are actively being conducted, the fire localization is still unfinished due to the high cost and low performance. This paper presents the modeling and simulation of the fire localization estimation using Bayesian estimation to determine the probabilistic location of the fire. To minimize the risk of fire accidents as well as the time and cost of preparing and executing live fire tests, a 40m × 40m-virtual space is created, where two ultraviolet sensors are simulated to rotate horizontally to collect ultraviolet signals. In addition, Bayesian estimation is executed to compute the probability of the fire location by considering both sensor errors and uncertainty under fire environments. For the validation of the proposed method, sixteen fires were simulated in different locations and evaluated by calculating the difference in distance between simulated and estimated fire locations. As a result, the proposed method demonstrates reliable outputs, showing that the error distribution tendency widens as the radial distance between the sensor and the fire increases.

• 한국건축시공학회지
• /
• 제3권2호
• /
• pp.141-148
• /
• 2003

This study deals with the estimation of material properties according to the construction condition for water-soluble rubberized asphalt waterproofing material of spray type. In this study, the waterproofing material property by the spray construction method is suggested by means of estimation its tensile performance and temperature dependency according to mix proportion ratio(4:1, 8:1), referenced viscosity and solid content (A:360cps, 76%, B:580cps, 79%, C:490cps, 70%), spray angle($30^{\circ}$, $45^{\circ}$, $60^{\circ}$), and spray distance(30cm, 50cm, 70cm). The result of testing are as follows. (1) The mix proportion ratio of principal agent and hardener is 4:1. (2) The viscosity referenced and solid content are 490cps and 70%. (3) The spray angle referenced is $45.^{\circ}$ (4) The distance referenced from concrete surface to spray gun is 40~50cm.

• 로봇학회논문지
• /
• 제11권1호
• /
• pp.33-40
• /
• 2016

In this paper, we propose a method for estimating the pose of the camera using a rectangle feature utilized for the visual SLAM. A warped rectangle feature as a quadrilateral in the image by the perspective transformation is reconstructed by the Coupled Line Camera algorithm. In order to fully reconstruct a rectangle in the real world coordinate, the distance between the features and the camera is needed. The distance in the real world coordinate can be measured by using a stereo camera. Using properties of the line camera, the physical size of the rectangle feature can be induced from the distance. The correspondence between the quadrilateral in the image and the rectangle in the real world coordinate can restore the relative pose between the camera and the feature through obtaining the homography. In order to evaluate the performance, we analyzed the result of proposed method with its reference pose in Gazebo robot simulator.

• 설비공학논문집
• /
• 제15권3호
• /
• pp.196-202
• /
• 2003

Two dimensional modeling was carried out to find the safety distance between the defrost heater and the plastic inner wall of domestic refrigerator Estimation was processed for the three cases; the estimation of plastic wall temperature (1) without any protection, (2) with an aluminum foil attached on a wall, and (3) with an aluminum shield installed between heater and wall. The former two cases are found to be dangerous during defrosting process, because the temperatures of inner wall reach above 80'C , which is the upper temperature limit of the wall material. The case with an aluminum shield is considered to be safe by maintaining the temperature of the wall in the range of 6$0^{\circ}C$ during defrosting process.

• 한국군사과학기술학회지
• /
• 제8권3호
• /
• pp.5-10
• /
• 2005

This paper presents the new in-flight alignment method using the flight distance of vehicle in order to improve the performance of the heading error estimation. In the proposed method, the Kalman filter having the difference between GPS and SDINS position as measurements is used for levelling of SDINS and heading error is estimated utilizing the flight distance information. It is shown in the simulation results that the in-flight method proposed in this paper has the high accuracy in heading error estimation and the heading error can be very quickly estimated at the high speed vehicle, compared with the existing method using the Kalman filter.

• 한국산학기술학회논문지
• /
• 제22권5호
• /
• pp.561-566
• /
• 2021

오차분포 추정을 위한 커널 사이즈는 오차확률밀도 사이의 유클리드 거리를 최소화 알고리즘의 가중치 갱신에 적합한 커널 사이즈가 될 수 없다. 이 논문에서는 MED 알고리즘의 수렴 성능 향상을 위해 적응적으로 커널 사이즈를 갱신하는 방법을 제안하였다. 제안한 방식은 MED 학습 알고리즘의 가중치 갱신을 위해 커널 사이즈에 대한 오차분산의 평균변화율을 도입하여 MED의 오차에 대한 평균전력이 감소하는 방향으로 커널 사이즈를 조절하도록 하였다. 제안된 적응 커널 추정법을 무선통신 채널의 왜곡 보상에 적용하여 학습 성능을 실험하고 그 효능을 밝혔다. 오차분산에 비례한 작은 값을 가지는 기존의 오차분포 추정 위한 최적 커널 사이즈와 달리, 제안한 방법에 의한 커널 사이즈는 MED 가중치 수렴을 위한 적절한 커널 사이즈로 수렴함을 보였다. 실험 결과로부터 제안한 방법이 MED 알고리즘의 커널 사이즈 설정에 따른 민감성을 크게 해결한 방법이라고 볼 수 있다.