• 한국측량학회지
• /
• 제38권3호
• /
• pp.223-235
• /
• 2020

본 연구에서는 무인항공기 영상 기반의 정밀농업(precision agricultural) 구현에 있어 핵심 데이터 중 하나인 수치표고모델의 표고를 보정하기 위한 수치표고모델 표고 보정 방법론을 제시한다. 먼저 정사영상에 방사보정을 수행한 다음 ExG (Excess Green)를 생성한다. ExG에 Otsu 기법을 적용하여 산출된 임계값을 기준으로 비식생지역을 추출한다. 이어서, 비식생지역의 위치에 대응되는 수치표고모델의 표고를 표고 보정을 위한 데이터인 EIFs(Elevation Invariant Features)로 추출한다. 추출된 EIFs 간 차이값을 기반으로 정규화된 Z-score를 산출하여 포함된 특이치를 제거한다. 그리고 선형회귀식을 구성하여 수치표고모델의 표고를 보정함으로써 지상기준점 데이터 없이 고품질의 수치표고모델을 제작한다. 총 10장의 수치표고모델을 활용하여 제안기법을 검증하기 위해 표고 보정 전과 후의 최대/최소값, 평균/표준편차를 비교분석하였다. 또한, 검사점을 선정하여 RMSE (Root Mean Square Error)를 산출한 결과, 정확도는 평균 RMSE 0.35m로 도출되었다. 이를 통해 지상기준점 데이터 없이 고품질의 수치표고모델을 제작할 수 있음을 확인하였다.

• 제어로봇시스템학회:학술대회논문집
• /
• 제어로봇시스템학회 2001년도 ICCAS
• /
• pp.62.2-62
• /
• 2001

In this paper, we present a new time-varying sliding mode system that guarantees stable error convergence. The previous papers have dealt with stability of the time-varying sliding mode system by point-wisely investigating the stability of time-invariant system every time. However, it may be unstable even though it guarantees time-invariant stability every time, We designed the time-varying sliding surface so that the resultant time-varying system on sliding mode may be Stable. The initial sliding surface is obtained so that shifting distance of the surface may be minimized with respect to an initial error, and the intercept is produced so that the surface may pass the initial error.

• Journal of information and communication convergence engineering
• /
• 제21권2호
• /
• pp.139-144
• /
• 2023

Engineers prefer deep neural networks (DNNs) for solving computer vision problems. However, DNNs pose two major problems. First, neural networks require large amounts of well-labeled data for training. Second, the covariate shift problem is common in computer vision problems. Domain adaptation has been proposed to mitigate this problem. Recent work on adversarial-learning-based unsupervised domain adaptation (UDA) has explained transferability and enabled the model to learn robust features. Despite this advantage, current methods do not guarantee the distinguishability of the latent space unless they consider class-aware information of the target domain. Furthermore, source and target examples alone cannot efficiently extract domain-invariant features from the encoded spaces. To alleviate the problems of existing UDA methods, we propose the mixup regularization in adversarial discriminative domain adaptation (ADDA) method. We validated the effectiveness and generality of the proposed method by performing experiments under three adaptation scenarios: MNIST to USPS, SVHN to MNIST, and MNIST to MNIST-M.

• 한국지능시스템학회:학술대회논문집
• /
• 한국퍼지및지능시스템학회 1995년도 추계학술대회 학술발표 논문집
• /
• pp.71-77
• /
• 1995

In this paper are proposed robust iterative learning control(ILC) algorithms for both linear continuous time-invariant system and linear discrete-time system. In contrast to conventional methods, the proposed learning algorithms are constructed based on both time domain performance and iteration-domain performance. The convergence of the proposed learning algorithms is proved. Also, it is shown that the proposed method has robustness in the presence of external disturbances and the convergence accuracy can be improved. A numerical external disturbances and the convergence accuracy can be improved. A numerical example is provided to show the effectiveness of the proposed algorithm.

• Journal of applied mathematics & informatics
• /
• 제30권3_4호
• /
• pp.365-380
• /
• 2012

Non-negative matrix factorization (NMF) is a very efficient method to explain the relationship between functions for finding basis information of multivariate nonnegative data. The multiplicative update (MU) algorithm is a popular approach to solve the NMF problem, but it fails to approach a stationary point and has inner iteration and zero divisor. So the elementwisely alternating projected gradient (eAPG) algorithm was proposed to overcome the defects. In this paper, we use the fact that the equilibrium point is stable to prove the convergence of the eAPG algorithm. By using a classic model, the equilibrium point is obtained and the invariant sets are constructed to guarantee the integrity of the stability. Finally, the convergence conditions of the eAPG algorithm are obtained, which can accelerate the convergence. In addition, the conditions, which satisfy that the non-zero equilibrium point exists and is stable, can cause that the algorithm converges to different values. Both of them are confirmed in the experiments. And we give the mathematical proof that the eAPG algorithm can reach the appointed precision at the least iterations compared to the MU algorithm. Thus, we theoretically illustrate the advantages of the eAPG algorithm.

• Journal of Ubiquitous Convergence Technology
• /
• 제2권2호
• /
• pp.105-110
• /
• 2008

The definition of the program control is introduced on the theory of the basis of the first integrals SDE system. That definition allows constructing the program control gives opportunity to stochastic system to remain on the given dynamic variety. The program control is considered in terms of dynamically invariant for stochastic process.

• Journal of applied mathematics & informatics
• /
• 제38권1_2호
• /
• pp.1-12
• /
• 2020

In this paper, we solve the difference equation $x_{n+1}={\frac{x_nx_{n-2}}{ax_n-bx_{n-2}}}$, n = 0, 1, …, where a and b are positive real numbers and the initial values x_-2, x_-1 and x₀ are real numbers. We also find invariant sets and discuss the global behavior of the solutions of aforementioned equation.

• 대한수학회지
• /
• 제33권4호
• /
• pp.983-992
• /
• 1996

Let p(x, dy) be a transition probability function on $(S, \rho)$, where S is a complete separable metric space. Then a Markov process $X_n$ which has p(x, dy) as its transition probability may be generated by random iterations of the form $X_{n+1} = f(X_n, \varepsilon_{n+1})$, where $\varepsilon_n$ is a sequence of independent and identically distributed random variables (See, e.g., Kifer(1986), Bhattacharya and Waymire(1990)).

• 대한전자공학회:학술대회논문집
• /
• 대한전자공학회 1998년도 하계종합학술대회논문집
• /
• pp.689-692
• /
• 1998

Matching two images is an essential step for many computer vision applications. A new approach to the scale and rotation invariant scene matching is presented. A set of andidate parameters are hypthesized by mapping the angular difference and a new distance measure to the hough space and by detecting maximally consistent points. The proposed method is shown to be much faster than the conventinal one where the relaxation process is repeated until convergence, while providing robust matching performance, without a priori information on the geometrical transformation parameters.

• 제어로봇시스템학회:학술대회논문집
• /
• 제어로봇시스템학회 1989년도 한국자동제어학술회의논문집; Seoul, Korea; 27-28 Oct. 1989
• /
• pp.160-165
• /
• 1989

This paper deals with the robust controller design of robotic manipulator to track a desired trajectory in spite of the presence of unmodelled dynamics in cause of nonlinearity and parameter uncertainty. The approach followed in this paper is based on model reference adaptive control technique and convergence on hyperstability theory but it does away with assumption that process is characterized by a linear model remaining time invariant during adaptation process. A computer simulation has been performed to demonstrate the performance of the designed control system in task coordinates for stanford manipulator with payload and disturbances.