• 한국인터넷방송통신학회논문지
• /
• 제20권1호
• /
• pp.225-229
• /
• 2020

이미지 인식 및 영상처리, 컴퓨터 비전 등의 분야에서 합성곱 인공신경망 (Convolutional Neural Network, CNN)은 다양하게 응용되고 탁월한 성능을 내고 있다. 본 논문에서는 CNN을 활용한 이미지 인식 시스템에서 인식률을 저하시키는 요인 중 하나인 이미지의 회전에 대한 해결책으로써 CNN 기반 이미지 회전 보정 알고리즘을 제안한다. 본 논문에서는 Leeds Sports Pose 데이터셋을 활용하여 이미지를 임의의 각도만큼 회전시킨 학습데이터로 인공지능 모델을 학습시켜 출력으로 회전된 각도를 추정하도록 실험을 진행하였다. 학습된 인공지능 모델을 100장의 테스트 데이터 이미지로 실험하여 mean absolute error (MAE) 성능지표를 기준으로 4.5951의 값을 얻었다.

• 한국공작기계학회논문집
• /
• 제14권3호
• /
• pp.54-59
• /
• 2005

The cylindrical object are widely used as mechanical parts in the manufacturing process. In order to handling those objects using a robot or an automated machine automatically, the pose of the object must be known. The pose can be described by two rotation angles; one angle about the x axis and the other about the y axis. In the many previous researches these angles were obtained by the computationally intensive algorithm, that is, fitting the data as a polynomial and doing pseudo inverse. So that, this method required high performance microprocessor. In this paper in order to avoid complex computation, a new method based on a genetic algorithm is proposed and analyzed through a series of simulations. This algorithm utilized the geometry of the cylindrical shape. The simulation results show that this method find the pose angles very well In most cases, but the computation time is randomly changed because the genetic algorithm is basically one of the random search method.

• 제어로봇시스템학회논문지
• /
• 제20권12호
• /
• pp.1284-1289
• /
• 2014

In this Paper, outdoor position estimation system was implemented using GPS (Global Positioning System) and INS (Inertial Navigation System). GPS position information has lots of errors by interference from obstacles and weather, the surrounding environment. To reduce these errors, multiple GPS system is used. Also, the Discrete Wavelet Transforms was applied to INS data for compensation of its error. In this paper, position estimation of the mobile robot in the straight line is conducted by EKF (Extended Kalman Filter). However, curve running position estimation is less accurate than straight line due to phase change in rotation. The curve is recognized through the rate of change in heading angle and the position estimation precision of the initial curve was improved by UPF (Unscented Particle Filter). In the case of UPF, if the number of particle is so many that big memory gets size is needed and processing speed becomes late. So, it only used the position estimation in the initial curve. Thereafter, the position of mobile robot in curve is estimated through switching from UPF to EKF again. Through the experiments, we verify the superiority of the system and make a conclusion.

• 정보보호학회논문지
• /
• 제24권4호
• /
• pp.607-616
• /
• 2014

현재 영상감시 시스템에서 얼굴 인식을 통한 사람의 신원 확인은 정면 얼굴이 아닌 관계로 매우 어려운 기술에 속한다. 일반적인 사람들의 얼굴 영상과 입력된 얼굴 영상을 비교하여 유사도를 파악하고 신원을 확인 하는 기술은 각도의 차이에 따라 정확도의 오차가 심해진다. 이런 문제를 해결하기 위해 본 논문에서는 POSIT을 사용하여 얼굴 포즈 측정을 하고, 추정된 각도를 이용하여 3D 얼굴 영상을 제작 후 매칭 하여 일반적인 정면 영상끼리의 매칭이 아닌 rotated face를 이용한 매칭을 해보기로 한다. 얼굴을 매칭 하는 데는 상용화된 얼굴인식 알고리즘을 사용하였다. 얼굴 포즈 추정은 $10^{\circ}$이내의 오차를 보였고, 얼굴인증 성능은 약 95% 정도임을 확인하였다.

• 대한전기학회논문지:시스템및제어부문D
• /
• 제54권1호
• /
• pp.26-32
• /
• 2005

Rotor position information is essential in the operation of the switched reluctance motor(SRM) drive for generation the phase current switching signals. When an incremental encoder is used as a rotor position sensor, the initial rotor position can not be detected. Some sensorless rotor position estimation methods also have the same problem. In these systems, to initially align the rotor, the forced alignment method has a delay and reverse rotation before the motor can start. Therefore it can not be acceptable for unidirectional drive systems. So the forced alignment method is not desirable in all drive systems and the research on the SRM drives should be directed to a system without rotor alignment. In this paper, a new detection method of initial rotor position using the rate of change of current is suggested. Firstly, di/dt versus θ/sub R/ reference table, which is the relation between the rate of change of current and rotor position, is generated and then the squared Euclidean distance method is used to estimate the rotor position based on the table. The simulated and experimental results are presented demonstrating the feasibility and accuracy of this method.

• Smart Structures and Systems
• /
• 제16권2호
• /
• pp.295-328
• /
• 2015

In recent years the monitoring of structural behavior through acquisition of vibrational data has become common practice. In addition, recent advances in sensor development have made the collection of diverse dynamic information feasible. Other than the commonly collected acceleration information, Global Position System (GPS) receivers and non-contact, optical techniques have also allowed for the synchronous collection of highly accurate displacement data. The fusion of this heterogeneous information is crucial for the successful monitoring and control of structural systems especially when aiming at real-time estimation. This task is not a straightforward one as measurements are inevitably corrupted with some percentage of noise, often leading to imprecise estimation. Quite commonly, the presence of noise in acceleration signals results in drifting estimates of displacement states, as a result of numerical integration. In this study, a new approach based on a time domain identification method, namely the Unscented Kalman Filter (UKF), is proposed for correcting the "drift effect" in displacement or rotation estimates in an online manner, i.e., on the fly as data is attained. The method relies on the introduction of artificial white noise (WN) observations into the filter equations, which is shown to achieve an online correction of the drift issue, thus yielding highly accurate motion data. The proposed approach is demonstrated for two cases; firstly, the illustrative example of a single degree of freedom linear oscillator is examined, where availability of acceleration measurements is exclusively assumed. Secondly, a field inspired implementation is presented for the torsional identification of a tall tower structure, where acceleration measurements are obtained at a high sampling rate and non-collocated GPS displacement measurements are assumed available at a lower sampling rate. A multi-rate Kalman Filter is incorporated into the analysis in order to successfully fuse data sampled at different rates.

• 대한전자공학회논문지SP
• /
• 제40권6호
• /
• pp.31-40
• /
• 2003

대부분의 얼굴인식 시스템은 현재 2차원 영상을 기반으로 많은 분야에 응용되고 있다. 그러나 2차원 얼굴인식 시스템은 심하게 변화된 얼굴 포즈에 강인한 얼굴인식이 매우 어렵다. 이에 얼굴 포즈 추정은 정면 영상이 아닐 경우 인식률 향상을 위한 필수적인 과정이라 할 수 있다. 그러므로, 본 논문은 3차원 얼굴인식을 위한 새로운 얼굴 포즈 추정 방식을 제안한다 먼저 3차원 거리(range) 영상이 입력될 때 얼굴 곡선에 기반한 자동 얼굴 특징점 추출 기법을 적용한다. 추출된 특징점을 바탕으로 오류 보상 특이치 분해를 적용 한 새로운 3차원 얼굴 포즈 추정 방식을 제안한다. 특이치 분해를 이용하여 초기 회전각을 획득한 후 존재하는 오류를 보다 세밀하게 보상한다. 제안 알고리즘은 정규화된 3차원 얼굴 공간에서 추출된 특징점의 기하학적 위치를 이용하여 수행된다. 또한 3차원 얼굴인식을 위하여 3차원 최근접 이웃 분류기를 이용한 데이터베이스내에서 후보 얼굴을 선택하는 방식을 제안한다. 실험 결과를 통해 다양한 얼굴 포즈에 대하여 제안 알고리즘의 효율성과 타당성을 검증하였다.

• Journal of Electrical Engineering and Technology
• /
• 제11권2호
• /
• pp.519-528
• /
• 2016

In order to achieve fast computational speed with good visual quality of output video, we propose a frequency domain based new fractal video compression scheme. Normalized cross correlation is used to find the structural self similar domain block for the input range block. To increase the searching speed, cross correlation is implemented in the frequency domain using FFT with one computational operation for all the domain blocks instead of individual block wise calculations. The encoding time is further minimized by applying rotation and reflection DFT properties to the IFFT of zero padded range blocks. The energy of overlap small size domain blocks is pre-computed for the entire reference frame and retaining the energies of the overlapped search window portion of previous adjacent block. Quadtree decompositions are obtained by using domain block motion compensated prediction error as a threshold to control the further partitions of the block. It provides a better level of adaption to the scene contents than fixed block size approach. The result shows that, on average, the proposed method can raise the encoding speed by 48.8 % and 90 % higher than NHEXS and CPM/NCIM algorithms respectively. The compression ratio and PSNR of the proposed method is increased by 15.41 and 0.89 dB higher than that of NHEXS on average. For low bit rate videos, the proposed algorithm achieve the high compression ratio above 120 with more than 31 dB PSNR.

• 한국항행학회논문지
• /
• 제16권5호
• /
• pp.738-743
• /
• 2012

INS/GPS 통합 항법 필터는 주로 칼만 필터를 이용하여 구성되고 있다. 그러나 낮은 성능의 관성 센서를 이용한 INS/GPS 통합 항법 필터의 경우 센서 오차 레벨 등의 문제로 인해 정확한 방위각 정보의 제공이 곤란하며, 이로 인해 통합 필터를 구성하는 칼만 필터의 추정 성능이나 안정성을 보장할 수 없게 된다. 본 논문에서는 칼만 필터와 파티클 필터의 결합된 형태의 국지선형 파티클 필터를 이용하여 초기 방위각 오차를 극복할 수 있는 방법을 제시하였다.

• Journal of Acupuncture Research
• /
• 제23권1호
• /
• pp.165-177
• /
• 2006

Objectives : 1. Clinical trial for the efficacy evaluation of Korean acupuncture techniques in treating frozen shoulder. 2. Development of the standard clinical guidelines of the acupuncture treatment for the frozen shoulder. 3. Development of the new clinical protocol for the acupuncture treatments. 4. Verification of the hypothesis that treating at both the remote and the nearby acupoints according to the meridian theory is more effective than treating at only the nearby acupoints. Methods : 1. Research designed as Single blind, Randomized, Sham acupuncture controlled clinical Trial. 2. Assignment of 86 patients to one of three groups treated at nearby acupoints(group A), remote & nearby acupoints(group B), and sham points(group C) respectively. 3. Trial conducted at KIOM CRC of Dunsan oriental medical hospital, Daejeon Univ. 4. Estimation of the recovery rate of the frozen shoulder in subjects aged over 40. 5. Efficacy evaluation using VAS, SPADI, ROM and Improvement rate. Results : 1. There was no significant difference in VAS among the three groups. 2. Pain related scores in SPADI of the group B were significantly lower than those of the group A. 3. There was no significant difference in disability related scores of SPADI among the three groups. 4. External rotation of upper arm in the group B was significantly improved in comparison with that in the group C. 5. Abduction of upper arm in group A was improved with weak statistical significance in comparison with that in the group C. Conclusion : Acupuncture at both the remote and the nearby acupoints according to the meridian theory is effective to improve external rotation of frozen shoulder, and acupuncture at the nearby acupoints is effective to improve adduction of frozen shoulder. However it is not clear that acupuncture treatment at both the remote and the nearby acupoints according to the meridian theory is more effective than treating at only the nearby acupoints in the treatment of frozen shoulder. Since our study was a short term trial, a long term trial for a more precise evaluation of acupuncture treatment for frozen shoulder will be needed in the future.