• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.8 no.2
• /
• pp.462-482
• /
• 2014

This paper proposes a new resampling detection method for images that detects whether an image has been resampled and recovers the corresponding resampling rate. The proposed method uses a given set of zeroing masks for various resampling factors to evaluate the convolution values of the input image with the zeroing masks. Improving upon our previous work, the proposed method detects more resampling factors by checking for some periodicity with an approximate detection mechanism. The experimental results demonstrate that the proposed method is effective and efficient.

• Journal of the Institute of Electronics Engineers of Korea SP
• /
• v.46 no.2
• /
• pp.78-88
• /
• 2009

In many face recognition problems, the number of available images is limited compared to the dimension of the input space which is usually equal to the number of pixels. This problem is called as the 'small sample size' problem and regularization methods are typically used to solve this problem in feature extraction methods such as LDA. By using regularization methods, the modified within class matrix becomes nonsingu1ar and LDA can be performed in its original form. However, in the process of adding a scaled version of the identity matrix to the original within scatter matrix, the scale factor should be set heuristically and the performance of the recognition system depends on highly the value of the scalar factor. By using the proposed resampling method, we can generate a set of images similar to but slightly different from the original image. With the increased number of images, the small sample size problem is alleviated and the classification performance increases. Unlike regularization method, the resampling method does not suffer from the heuristic setting of the parameter producing better performance.

• Journal of Korean Society for Quality Management
• /
• v.26 no.4
• /
• pp.250-264
• /
• 1998

The purpose of the thesis is to develop simulation program for forecasting of optical connector. So we can achieve the time and the money saving for making the optical connector. Optical performance (insertion loss) of optical connector mainly relies on 3 misalignment factors-ferrule factor due to mis-manufacture from design, auto-centering effect that is fiber behavior phenomena between hole and fiber, fiber misalignment factor. Simulation use experimental data with auto-centering effect and fiber factor and use pseudo data with ferrule through random number generation because it is developing stage. In this study we a, pp.y kernel density estimation method with experimental data in order to know whether it belong to or not specific parametric distribution family. And we simulate to forecast insertion loss of optical multifiber connector under specific design model using nonparametric bootstrap resampling data and parametric pseudo samples from uniform distribution. We obtain the tolerance specifications of misalignment factors satisfying not exceed in maximum 1.0dB and choose optimal hole diameter.

• Communications for Statistical Applications and Methods
• /
• v.19 no.3
• /
• pp.495-500
• /
• 2012

We consider a Bayesian test of independence in a two-way contingency table that has some zero cells. To do this, we take a three-stage hierarchical Bayesian model under each hypothesis. For prior, we use Dirichlet density to model the marginal cell and each cell probabilities. Our method does not require complicated computation such as a Metropolis-Hastings algorithm to draw samples from each posterior density of parameters. We draw samples using a Gibbs sampler with a grid method. For complicated posterior formulas, we apply the Monte-Carlo integration and the sampling important resampling algorithm. We compare the values of the Bayes factor with the results of a chi-square test and the likelihood ratio test.

• Journal of Korea Multimedia Society
• /
• v.7 no.1
• /
• pp.80-90
• /
• 2004

Watermarking algorithm for 3D mesh model using EGI distribution of each patch is proposed. The proposed algorithm divides a 3D mesh model into 6 patches to have the robustness against the partial geometric deformation. Plus, it uses EGI distributions as the consistent factor that has the robustness against the topological deformation. To satisfy both geometric and topological deformation, the same watermark bits for each subdivided patch are embedded by changing the mesh normal vectors. Moreover, the proposed algorithm does not need not only the original mesh model but also the resampling process to extract the watermark. Experimental results verify that the proposed algorithm is imperceptible and robust against geometrical and topological attacks.

• Proceedings of the Korea Water Resources Association Conference
• /
• 2021.06a
• /
• pp.397-397
• /
• 2021

본 연구의 목적은 비매개변수적 리샘플링 기법을 이용하여 농업용 저수지 유입 설계홍수량의 구간을 추정하는 기법을 제안하는 데 있다. 본 연구는 설계홍수량을 점 추정하여 안전계수(safety factor)를 적용하는 기존 방법에 대한 대안을 제시하고자 한다. 설계홍수량의 구간 추정을 수행하기 위해 부트스트랩 기법(bootstrap technique)을 사용하였다. 부트스트랩 기법을 이용하여 95% 신뢰수준에 해당하는 신뢰구간을 추정하였다. 본 연구의 공간적인 범위는 남한의 30개 농업용 저수지이며, 시간적인 범위는 과거 기간(2015s: 1986-2015)과 미래기간(2040s: 2011-2040, 2070s: 2041-2070, 2100s: 2071-2100)을 설정하였다. 본 연구에서는 200년 빈도, 24시간 지속기간을 대표적인 결과로 선정하여 분석하였다. 빈도분석은 GEV 분포를 사용하였고, L-moment 방법을 이용하여 매개변수를 추정하였다. 설계홍수량은 HEC-1 모형을 이용하여 산정하였다. 최종적으로 설계홍수량 구간 추정한 결과를 기존의 점 추정한 뒤 안전계수를 적용한 기존 방법과 비교하였다. 97.5th BCa percentile 기준으로 상대적인 변화를 비교해보면, 미래로 갈수록 구간 추정으로 산정한 설계홍수량이 점차 증가하는 것으로 도출되었다. 한강 및 금강 유역에 위치한 농업용 저수지의 설계홍수량이 낙동강 유역에 비해 상대적으로 큰 변화를 보여주었다. 몇몇 농업용 저수지에 대해서 2040s 기간에 다소 감소하기도 하였으나 2070s 기간 이후에 다시 증가하는 결과를 보여주었다. 낙동강 유역의 위치는 농업용 저수지의 설계홍수량은 미래로 갈수록 크게 증가하지 않는 경향을 보여주었다. 본 연구는 설계홍수량을 추정하는 데 있어 결정론적인 방법에서 더 나아가 자료의 통계적인 특성을 고려하여 구간 추정을 수행하는 방법론을 제공할 수 있을 것으로 사료된다.