• 응용통계연구
• /
• 제25권4호
• /
• pp.669-680
• /
• 2012

Noise removal of document images is a necessary step during preprocessing to recognize characters effectively because it has influences greatly on processing speed and performance for character recognition. We have considered using the spatial filters such as traditional mean filters and Gaussian filters, and wavelet transformed based methods for noise deduction in natural images. However, these methods are not effective for the noise removal of document images. In this paper, we present noise removal of document images using support vector regression. The proposed approach consists of two steps which are SVR training step and SVR test step. We construct an optimal prediction model using grid search with cross-validation in SVR training step, and then apply it to noisy images to remove noises in test step. We evaluate our SVR based method both quantitatively and qualitatively for noise removal in Korean, English and Chinese character documents, and compare it to some existing methods. Experimental results indicate that the proposed method is more effective and can get satisfactory removal results.

• Nuclear Engineering and Technology
• /
• 제41권9호
• /
• pp.1181-1190
• /
• 2009

Knowing more about the Local Power Density (LPD) at the hottest part of a nuclear reactor core can provide more important information than knowledge of the LPD at any other position. The LPD at the hottest part needs to be estimated accurately in order to prevent the fuel rod from melting in a nuclear reactor. Support Vector Machines (SVMs) have successfully been applied in classification and regression problems. Therefore, in this paper, the power peaking factor, which is defined as the highest LPD to the average power density in a reactor core, was estimated by SVMs which use numerous measured signals of the reactor coolant system. The SVM models were developed by using a training data set and validated by an independent test data set. The SVM models' uncertainty was analyzed by using 100 sampled training data sets and verification data sets. The prediction intervals were very small, which means that the predicted values were very accurate. The predicted values were then applied to the first fuel cycle of the Yonggwang Nuclear Power Plant Unit 3. The root mean squared error was approximately 0.15%, which is accurate enough for use in LPD monitoring and for core protection that uses LPD estimation.

• Nuclear Engineering and Technology
• /
• 제40권1호
• /
• pp.69-76
• /
• 2008

Most pressurized water reactors use Venturi flow meters to measure the feedwater flow rate. However, fouling phenomena, which allow corrosion products to accumulate and increase the differential pressure across the Venturi flow meter, can result in an overestimation of the flow rate. In this study, a soft-sensing model based on fuzzy support vector regression was developed to enable accurate on-line prediction of the feedwater flow rate. The available data was divided into two groups by fuzzy c means clustering in order to reduce the training time. The data for training the soft-sensing model was selected from each data group with the aid of a subtractive clustering scheme because informative data increases the learning effect. The proposed soft-sensing model was confirmed with the real plant data of Yonggwang Nuclear Power Plant Unit 3. The root mean square error and relative maximum error of the model were quite small. Hence, this model can be used to validate and monitor existing hardware feedwater flow meters.

• 제어로봇시스템학회논문지
• /
• 제12권4호
• /
• pp.358-363
• /
• 2006

This paper demonstrates a new non-contact electronic joystick using single hall sensor which detects a horizontal vector of the magnetic field. Furthermore, in this paper, it is mathematically modeled that nonlinear characteristics between the output of hall sensor and the movement of joystick bar. The dynamic horizontal vector of magnetic flux is detected by the hall sensor while a permanent magnet is rotated with the joystick bar, which has two dimension detecting area. Using the nonlinear adjustment equations, the output signals of hall sensor have been linearized to give higher accuracy in the two dimension movement. Finally, through the real experiments, it is showed that the single hall sensor structure mechanism is superior to the dual sensor structure in sensing the two-dimensional motion without offset.

• 대한기계학회:학술대회논문집
• /
• 대한기계학회 2001년도 춘계학술대회논문집E
• /
• pp.702-707
• /
• 2001

The characteristics of flow in dividing regions are precise, therefore their classification is very important not only in industry but also in hydrodynamics. By now, many studies of flow in dividing regions have been performed, but flow characteristics that use visualization in dividing regions have not been studied. The present study of the PIV and the CFD exhibit average velocity distributions, kinetic energy distributions and total pressure distributions etc of the total flow field due to the development of the accurate visualization optical laser and of optical equipment. Also, PIV is accurate with the flows characteristics of the dividing region as continuous analysis is done using input equipment. The study analyzes average velocity vector field, average kinetic energy, x-axis stress distributions, average and total pressure distributions of dividing regions with flow for visualization of the PIV and the CFD measurement in a dividing rectangular duct.

• ETRI Journal
• /
• 제32권5호
• /
• pp.766-773
• /
• 2010

Support vector machine (SVM) active learning plays a key role in the interactive content-based image retrieval (CBIR) community. However, the regular SVM active learning is challenged by what we call "the small example problem" and "the asymmetric distribution problem." This paper attempts to integrate the merits of semi-supervised learning, ensemble learning, and active learning into the interactive CBIR. Concretely, unlabeled images are exploited to facilitate boosting by helping augment the diversity among base SVM classifiers, and then the learned ensemble model is used to identify the most informative images for active learning. In particular, a bias-weighting mechanism is developed to guide the ensemble model to pay more attention on positive images than negative images. Experiments on 5000 Corel images show that the proposed method yields better retrieval performance by an amount of 0.16 in mean average precision compared to regular SVM active learning, which is more effective than some existing improved variants of SVM active learning.

• 한국식품영양학회지
• /
• 제9권2호
• /
• pp.123-136
• /
• 1996

IPSKCDNA gene(1.8 kbp) encoding rat brain IP3K enzyme contained Not I restric site in open reading frame. The Not I sequence, GCGGCCGC, was converted to GCAGCCGC by site-directed mutagenesis. The mutated IP3KcDNA was digested with EcoR I and ligated with EcoR I-restricted psp72·Not2 vector. The resulting psp72 · Not2-IP3KCDNA was digested with the Not I restriction enzyme and then subcloned into the Not I -digested PZIP · NeoSV(X) mammalian expression vector. The PZIP · NeoSV(X) -IPSKCDNA was transfected into CCL39 hamster lung fibroblast cells. The efficiency of the expressed IPSKCDNA gene was significantly higher than expected generally, not only a mean 5-fold increase in the amount of enzyme, but also 16-fold increase in enzyme activity from tractsfected CCL39 cells by the method of Western blot using anti-lP3K antibodies. Both distribution of IPSK in various rat tissues and biochemical properties were discussed.

• 대한수학회보
• /
• 제50권3호
• /
• pp.935-949
• /
• 2013

In this paper, we study surfaces in $\mathb{E}^3$ whose Gauss map G satisfies the equation ${\Box}G=f(G+C)$ for a smooth function $f$ and a constant vector C, where ${\Box}$ stands for the Cheng-Yau operator. We focus on surfaces with constant Gaussian curvature, constant mean curvature and constant principal curvature with such a property. We obtain some classification and characterization theorems for these kinds of surfaces. Finally, we give a characterization of surfaces whose Gauss map G satisfies the equation ${\Box}G={\lambda}(G+C)$ for a constant ${\lambda}$ and a constant vector C.

• 대한전자공학회논문지
• /
• 제26권10호
• /
• pp.1561-1570
• /
• 1989

In this paper, we propose an image coding scheme which combines the Laplacian pyramid structure and a hierarchical finite state classified vector quantizer in the DCT domain, namely FSDCT-CTQ. First, an optimal bit allocation problem for fixed rates DCT-CVQ on the Laplacian pyramid structure is described. In an asymptotic case, with an optimal bit allocation, a coding gain over scalar quantization of each Laplacian plane is derived. Second, it is experimentallhy shown that the Laplacian pyramid structure provides a considerable codng gain in the sense of total MMSE (minimum mean squared error). Finally, we propose an FS-DCT-CVQ which exploits the hierarchicla correlation between the Laplacian planes. Simulation results on real images show that the proposed coding scheme can reconstruct an image with 30.33 dB at 0.192 bpp, 32.45 dB at 0.385 bpp, respectively.

• 한국지능시스템학회논문지
• /
• 제11권7호
• /
• pp.579-583
• /
• 2001

In the area of data clustering in high dimensional space, one of the difficulties is the time-consuming process for computing vector similarities. It becomes worse in the case of the agglomerative algorithm with the group-average link and mean centroid method, because the cluster similarity must be recomputed whenever the cluster center moves after the merging step. As a solution of this problem, we present an incremental method of similarity computation, which substitutes the scalar calculation for the time-consuming calculation of vector similarity with several measures such as the squared distance, inner product, cosine, and minimum variance. Experimental results show that it makes clustering speed significantly fast for very high dimensional data.