• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.23 no.6
• /
• pp.675-681
• /
• 2019

Noise removal is a pre-requisite procedure in image processing, and various methods have been studied depending on the type of noise and the environment of the image. However, for image processing with high-frequency components, conventional additive white Gaussian noise (AWGN) removal techniques are rather lacking in performance because of the blurring phenomenon induced thereby. In this paper, we propose an algorithm to minimize the blurring in AWGN removal processes. The proposed algorithm sets the high-frequency and the low-frequency component filters, respectively, depending on the pixel properties in the mask, consequently calculating the output of each filter with the addition or subtraction of the input image to the reference. The final output image is obtained by adding the weighted data calculated using the standard deviations and the Gaussian distribution with the output of the two filters. The proposed algorithm shows improved AWGN removal performance compared to the existing method, which was verified by simulation.

• Journal of Institute of Control, Robotics and Systems
• /
• v.20 no.11
• /
• pp.1098-1102
• /
• 2014

Localization of aerial vehicles and map building of flight environments are key technologies for the autonomous flight of small UAVs. In outdoor environments, an unmanned aircraft can easily use a GPS (Global Positioning System) for its localization with acceptable accuracy. However, as the GPS is not available for use in indoor environments, the development of a SLAM (Simultaneous Localization and Mapping) system that is suitable for small UAVs is therefore needed. In this paper, we suggest a vision-based SLAM system that uses vision sensors and an AHRS (Attitude Heading Reference System) sensor. Feature points in images captured from the vision sensor are obtained by using GPU (Graphics Process Unit) based SIFT (Scale-invariant Feature Transform) algorithm. Those feature points are then combined with attitude information obtained from the AHRS to estimate the position of the small UAV. Based on the location information and color distribution, a Gaussian process model is generated, which could be a map. The experimental results show that the position of a small unmanned aircraft is estimated properly and the map of the environment is constructed by using the proposed method. Finally, the reliability of the proposed method is verified by comparing the difference between the estimated values and the actual values.

• Communications for Statistical Applications and Methods
• /
• v.24 no.4
• /
• pp.383-396
• /
• 2017

The model in our approach assumes that computer responses are a realization of a Gaussian processes superimposed on a regression model called a Gaussian process regression model (GPRM). Selecting a subset of variables or building a good reduced model in classical regression is an important process to identify variables influential to responses and for further analysis such as prediction or classification. One reason to select some variables in the prediction aspect is to prevent the over-fitting or under-fitting to data. The same reasoning and approach can be applicable to GPRM. However, only a few works on the variable selection in GPRM were done. In this paper, we propose a new algorithm to build a good prediction model among some GPRMs. It is a post-work of the algorithm that includes the Welch method suggested by previous researchers. The proposed algorithms select some non-zero regression coefficients (${\beta}^{\prime}s$) using forward and backward methods along with the Lasso guided approach. During this process, the fixed were covariance parameters (${\theta}^{\prime}s$) that were pre-selected by the Welch algorithm. We illustrated the superiority of our proposed models over the Welch method and non-selection models using four test functions and one real data example. Future extensions are also discussed.

• The Journal of Korea Robotics Society
• /
• v.12 no.4
• /
• pp.402-410
• /
• 2017

Acoustic based localization is essential to operate autonomous robotic systems in underwater environment where the use of sensorial data is limited. This paper proposes a localization method using artificial underwater acoustic sources. The proposed method acquires directional angles of acoustic sources using time difference of arrivals of two hydrophones. For this purpose, a probabilistic approach is used for accurate estimation of the time delay. Then, Gaussian sum filter based SLAM technique is used to localize both acoustic sources and underwater vehicle. It is performed by using bearing of acoustic sources as measurement and inertial sensors as prediction model. The proposed method can handle directional ambiguity of time difference based source localization by generating Gaussian models corresponding to possible locations of both front and back sides. Through these processes, the proposed method can provide reliable localization method for underwater vehicles without any prior information of source locations. The performance of the proposed method is verified by experimental results conducted in a real sea environment.

• Journal of the Korean Mathematical Society
• /
• v.58 no.1
• /
• pp.237-264
• /
• 2021

In this work the stationary bootstrap of Politis and Romano [27] is applied to the empirical distribution function of stationary and associated random variables. A weak convergence theorem for the stationary bootstrap empirical processes of associated sequences is established with its limiting to a Gaussian process almost surely, conditionally on the stationary observations. The weak convergence result is proved by means of a random central limit theorem on geometrically distributed random block size of the stationary bootstrap procedure. As its statistical applications, stationary bootstrap quantiles and stationary bootstrap mean residual life process are discussed. Our results extend the existing ones of Peligrad [25] who dealt with the weak convergence of non-random blockwise empirical processes of associated sequences as well as of Shao and Yu [35] who obtained the weak convergence of the mean residual life process in reliability theory as an application of the association.

• KSCE Journal of Civil and Environmental Engineering Research
• /
• v.2 no.3
• /
• pp.43-56
• /
• 1982

A recursive algorithms for prediction of streamflows by Kalman filtering theory and Self-tuning predictor based on the state space description of the dynamic systems have been studied and the applicabilities of the algorithms to the rainfall-runoff processes have been investigated. For the representation of the dynamics of the processes, a low-order ARMA process has been taken as the linear discrete time system with white Gaussian disturbances. The state vector in the prediction model formulated by a random walk process. The model structures have been determined by a statistical analysis for residuals of the observed and predicted streamflows. For the verification of the prediction algorithms developed here, the observed historical data of the hourly rainfall and streamflows were used. The numerical studies shows that Kalman filtering theory has better performance than the Self-tuning predictor for system identification and prediction in rainfall-runoff processes.

• Journal of the Korean Institute of Telematics and Electronics B
• /
• v.30B no.9
• /
• pp.76-83
• /
• 1993

A morphological filtering algorithm using directional information is presented. Directional filtering technique is effective in reducing noises and preserving edges. The proposed directional filtering is composed of two stage filtering processes. The opening and closing operations in the lst stage are performed for the pixels is aligned to the vertical, horizontal, and two diagonal directions, respectively. The opening operation supresses the positive impulse noises, while the closing operation the negative ones. Then, each directional result and their average value are filtered by the opening or closing operations in the 2nd stage. The averaging operation diminishes the effects of Gaussian noises in the homogeneous regions. Thus, the morphological operation in the 1 st stageremoves the impulse noises and in 2nd stage reduces. Gaussian ones. The experimental results show that the proposed filtering is superior to the existing nonlinear filtering in the aspects of the subjective quality. Also, the morphological filtering method reduces the computational loads.

• Journal of the Korean Statistical Society
• /
• v.12 no.1
• /
• pp.10-17
• /
• 1983

Let $X^{(n)}_j, j=1,2,\cdots,n, n=1,2,\cdots$ be a triangular array of random variables which arise naturally in a study of ferromagnetism in statistical mechanics. This paper presents weak convergence of random function $W_n(t)$, an appropriately normalized partial sum process based on $S^{(n)}_n = X^{(n)}_i+\cdot+X^{(n)}_n$. The limiting process W(t) is shown to be Gaussian when weak dependence exists. At the critical point where the change form weak to strong dependence takes place, W(t) turns out to be non-Gaussian. Our results are direct extensions of work by Ellis and Newmam (1978). An example is considered and the relation of these results to critical phenomena is briefly explained.

• Journal of information and communication convergence engineering
• /
• v.9 no.3
• /
• pp.336-340
• /
• 2011

Many approaches to image restoration are aimed at removing either gauss or impulse noise. This is because both types of degradation processes are distinct in nature, and hence they are easier to manage when considered separately. Nevertheless, it is possible to find them operating on the same image, which produces a hard damage. This happens when an image, already contaminated by Gaussian noise in the image acquisition procedure, undergoes impulsive corruption during its digital transmission. Here we proposed an algorithm first judge the type of the noise according to the difference values of pixel's neighborhood region and impulse noise's characteristic. Then removes the gauss noise by modified weighted mean filter and removes the impulse noise by modified nonlinear filter. The result of computer simulation on test images indicates that the proposed method is superior to traditional filtering algorithms. The proposed method can not only remove mixed noise effectively, but also preserve image details.

• The Transactions of The Korean Institute of Electrical Engineers
• /
• v.62 no.7
• /
• pp.1014-1019
• /
• 2013

Signal processing techniques based on fractional order calculus have been successfully applied in analyzing heavy-tailed non-Gaussian signals. It was found that the surface EMG signals from the muscles having nuero-muscular disease are best modeled by using the heavy-tailed non-gaussian random processes. In this regard, this paper describes an application of digital fractional order lowpass differentiators(FOLPD, weighted FOLPD) based on the fractional order calculus in detecting peaks of surface EMG signal. The performances of the FOLPD and WFOLPD are analyzed based on different filter length and varying MUAP wave shape from recorded and simulated surface EMG signals. As a results, the WFOLPD showed better SNR improving factors than the existing WLPD and to be more robust under the various surface EMG signals.