• 전자공학회논문지
• /
• 제50권4호
• /
• pp.182-188
• /
• 2013

본 논문에서는 정지하고 있는 배경에서 움직이는 목표물을 식별하기 위해 PDR(pulse doppler radar)을 이용하여 수집한 복소수 신호를 처리하는 복소수 SVM(support vector machine)을 제안한다. SVM은 패턴인식 분야에서 널리 이용되나 분류에 이용되는 특징이 대부분 실수 데이터이다. 제안된 복소수 SVM은 실수 데이터, 허수 데이터 정보와 실수부와 허수부 사이의 교차 정보를 모두 이용하여 이동하는 목표물의 분류를 수행한다. 복소수 SVM을 설계하기 위해 최적화 조건 적용 시 실수축과 허수축에 대한 슬랙변수를 고려하였고, 복소수 데이터에 대한 KKT(Karush-Kuhn-Tucker) 조건을 이용하였다. 또한 복소수 거리를 이용한 RBF(radial basis function)를 커널함수로 적용하였다. 제안된 복소수 SVM의 성능을 평가하기 위해 PDR 센서로 수집된 복소 데이터를 기존의 SVM과 복소수 SVM을 이용하여 분류한 결과 기존의 SVM에 비해 복소수 SVM의 식별결과가 개와 사람 각각 8%, 10% 향상되었다.

• 전자공학회논문지
• /
• 제52권8호
• /
• pp.117-125
• /
• 2015

본 논문은 '좌', '우' 방향 제어를 위해 취득된 EEG(Electroencephalogram) 신호 기반 분류 알고리즘과 EEG 센서, Labview, DAQ, Matlab, 주행로봇으로 구성된 방향 제어 시스템을 제안한다. 제안된 알고리즘은 DWT(Discrete Wavelet Transform)로 추출된 주파수대역 정보를 특징으로 이용하며, Fishers score를 이용하여 변별력이 높은 주파수 대역의 특징을 선별한다. 또한, SVM (Support Vector Machine)을 이용하여 분류 성능이 최고가 되는 특징벡터의 조합을 제안하고, 잘못된 판정에 의한 오동작을 방지하기 위한 MLD(Maximum Likelihood Decision) 기반의 판정보류 알고리즘도 제안한다. 제안된 알고리즘에 의해 선택된 4개의 특징벡터는 국제 표준 전극 배치법에 따른 P8 채널의 d2(16-32Hz), d5(2-4Hz) 주파수 대역의 전압의 절대 값 평균과 표준편차이다. SVM 분류기로 실험한 결과 98.75%의 정확도와 1.25%의 오류율 성능을 보였다. 또한, 오류 확률 70%를 판정 보류로 규정할 경우, 제안된 알고리즘은 인식률 95.63%의 정확도와 오류율 0%을 보였다.

• 대한임베디드공학회논문지
• /
• 제3권2호
• /
• pp.119-125
• /
• 2008

This paper presents the polygon-based Q-leaning and Cascade Support Vector Machine algorithm for object search with multiple robots. We organized an experimental environment with ten mobile robots, twenty five obstacles, and an object, and then we sent the robots to a hallway, where some obstacles were lying about, to search for a hidden object. In experiment, we used four different control methods: a random search, a fusion model with Distance-based action making (DBAM) and Area-based action making (ABAM) process to determine the next action of the robots, and hexagon-based Q-learning and dodecagon-based Q-learning and Cascade SVM to enhance the fusion model with DBAM and ABAM process.

• 대한전자공학회:학술대회논문집
• /
• 대한전자공학회 2003년도 신호처리소사이어티 추계학술대회 논문집
• /
• pp.557-560
• /
• 2003

Voice is one of the promising biometrics because it is one of the most convenient ways human would distinguish someone from others. The target of speaker verification is to divide the client from imposters. Support Vector Machine(SVM) is in the limelight as a binary classifier, so it can work well in speaker verification. In this paper, we combined SVM with genetic algorithm(GA) to reduce the dimensionality of input feature. Experiments were conducted with Korean connected digit database using different feature dimensions. The verification accuracy of SVM with GA is slightly lower than that of SVM, but the proposed algorithm has greater strength in the memory limited systems.

• 전기학회논문지
• /
• 제61권9호
• /
• pp.1336-1339
• /
• 2012

In this paper, we focus on solving the classification problem by using semisupervised learning strategy. Traditional classifiers are constructed based on labeled data in supervised learning. Labeled data, however, are often difficult, expensive or time consuming to obtain, as they require the efforts of experienced human annotators. Unlabeled data are significantly easier to obtain without human efforts. Thus, we use AdaBoost algorithm with SVM-KNN classifier to apply semisupervised learning problem and improve the classifier performance. Experimental results on both artificial and UCI data sets show that the proposed methodology can reduce the error rate.

• 정보보호학회논문지
• /
• 제21권6호
• /
• pp.161-169
• /
• 2011

최근 지능형 비디오 감시 시스템은 영상 분석 및 인식기술 등의 보다 진화된 기술 개발을 요구하고 있다. 특히, 비디오 영상에서 객체를 식별하기 위하여 Support Vector Machine(SVM)과 같은 기계학습 알고리즘이 이용된다. 그러나 SVM은 대용량의 데이터를 학습시키기 위하여 많은 계산량이 필요하기 때문에 수행시간을 효율적으로 감소시키기 위하여 병렬처리 기법을 적용할 필요가 있다. 본 논문에서는, 최근 사용이 증가하고 있는 멀티코어 프로세서를 활용한 SVM 학습의 병렬처리 방법을 제안한다. 4-코어 프로세서를 이용한 실험 결과, 제안 방법은 SVM 학습의 순차처리 방법과 비교하여 2.5배 정도 수행시간이 감소됨을 확인하였다.

• 제어로봇시스템학회논문지
• /
• 제18권6호
• /
• pp.540-545
• /
• 2012

In this paper, we propose a fault diagnosis method using artificial neural network and SVM (Support Vector Machine) to detect and isolate faults in the nonlinear systems. The proposed algorithm consists of two main parts: fault detection through threshold testing using a artificial neural network and fault isolation by SVM fault classifier. In the proposed method a fault is detected when the errors between the actual system output and the artificial neural network nominal system output cross a predetermined threshold. Once a fault in the nonlinear system is detected the SVM fault classifier isolates the fault. The computer simulation results demonstrate the effectiveness of the proposed SVM and artificial neural network based fault diagnosis method.

• 한국소음진동공학회:학술대회논문집
• /
• 한국소음진동공학회 2003년도 춘계학술대회논문집
• /
• pp.932-937
• /
• 2003

When oil pressure of damper is lost or relative stiffness of spring drops in vibration system, it can be fatally dangerous situation. A fault diagnosis method for vibration system using Support Vector Machine(SVM)is suggested in the paper. SVM is used to classify input data or applied to function regression. System status can be classified by judging input data based on optimal separable hyperplane obtained using SVM which learns normal and abnormal status. It is learned from the relationship of system state variables in term of spring, mass and damper. Normal and abnormal status are learned using phase plane as in put space, then the learned SVM is used to construct algorithm to predict the system status quantitatively

• 대한임베디드공학회논문지
• /
• 제11권6호
• /
• pp.361-367
• /
• 2016

In this paper, we propose an improved single view metrology (SVM) algorithm to accurately measure the height of objects. In order to accurately measure the size of objects, vanishing points have to be correctly estimated. There are two methods to estimate vanishing points. First, the user has to choose some horizontal and vertical lines in real world. Then, the user finds the cross points of the lines. Second, the user can obtain the vanishing points by using software algorithm such as [6-9]. In the former method, the user has to choose the lines manually to obtain accurate vanishing points. On the other hand, the latter method uses software algorithm to automatically obtain vanishing points. In this paper, we apply image resizing and edge sharpening as a pre-processing to the algorithm in order to improve performance. The estimated vanishing points algorithm create four vanishing point candidates: two points are horizontal candidates and the other two points are vertical candidates. However, a common image has two horizontal vanishing points and one vertical vanishing point. Thus, we eliminate a vertical vanishing point candidate by analyzing the histogram of angle distribution of vanishing point candidates. Experimental results show that the proposed algorithm outperforms conventional methods, [6] and [7]. In addition, the algorithm obtains similar performance with manual method with less than 5% of the measurement error.

• International Journal of Fuzzy Logic and Intelligent Systems
• /
• 제7권2호
• /
• pp.138-142
• /
• 2007

This paper presents an algorithm which extends the rapidly-exploring random tree (RRT) framework to deal with change of the task environments. This algorithm called the Retrieval RRT Strategy (RRS) combines a support vector machine (SVM) and RRT and plans the robot motion in the presence of the change of the surrounding environment. This algorithm consists of two levels. At the first level, the SVM is built and selects a proper path from the bank of RRTs for a given environment. At the second level, a real path is planned by the RRT planners for the: given environment. The suggested method is applied to the control of $KUKA^{TM}$, a commercial 6 DOF robot manipulator, and its feasibility and efficiency are demonstrated via the cosimulatation of $MatLab^{TM}\;and\;RecurDyn^{TM}$.