• 제어로봇시스템학회논문지
• /
• 제14권5호
• /
• pp.415-419
• /
• 2008

Detecting and correcting defects on LCD glasses early in the manufacturing process becomes important for panel makers to reduce the manufacturing costs and to improve productivity. Many attempts have been made and were successfully applied to detect and identify simple defects such as scratches, dents, and foreign objects on glasses. However, it is still difficult to robustly detect low-contrast defect region, called Mura or blemish area on glasses. Typically, these defect areas are roughly defined as relatively large, several millimeters of diameter, and relatively dark and/or bright region of low Signal-to-Noise Ratio (SNR) against background of low-frequency signal. The aim of this article is to present a robust algorithm to segment these blemish defects. Early 90's, a highly robust estimator, known as the Model-Fitting (MF) estimator was developed by X. Zhuang et. al. and have been successfully used in many computer vision application. Compared to the conventional Least-Square (LS) estimator the MF estimator can successfully estimate model parameters from a dataset of contaminated Gaussian mixture. Such a noise model is defined as a regular white Gaussian noise model with probability $1-\varepsilon$ plus an outlier process with probability $varepsilon$. In the sense of robust estimation, the blemish defect in images can be considered as being a group of outliers in the process of estimating image background model parameters. The algorithm developed in this paper uses a modified MF estimator to robustly estimate the background model and as a by-product to segment the blemish defects, the outliers.

• Journal of information and communication convergence engineering
• /
• 제9권6호
• /
• pp.666-670
• /
• 2011

In this paper, we propose a stable active contour based tracking method which utilizes the bimodal segmentation technique to obtain a background color diminished image frame. The proposed method overcomes the drawback of the Mansouri model which is liable to fall into a local minimum state when colors appear in the background that are similar to the target colors. The Mansouri model has been a foundation for active contour based tracking methods, since it is derived from a probability based interpretation. By stabilizing the model with the proposed speed function, the proposed model opens the way to extend probability based active contour tracking for practical applications.

• 한국정보통신학회논문지
• /
• 제4권1호
• /
• pp.123-130
• /
• 2000

본 논문은 고정된 카메라에서 얻어진 연속 영상으로부터 능동 윤곽선 모델을 이용하여 이동 물체의 윤곽선을 추출하는 방법을 제안한다. 주위 환경 변화에 강인한 처리를 위해 적응 배경 모델을 사용하였다. 물체 분할 모델은 얻어진 배경 영상과 현재 영상의 차영상으로부터 국부 영상의 임계값 이상의 화소를 찾아 연결한 영역을 분할하며, 형태학적 필터에 의하여 이동 물체의 경계 부분에서 발생하는 잡음을 제거하였다 분할된 이동 물체 윤곽선은 능동 윤곽선 모델을 이용하여 보다 정확한 이동 물체의 경계를 추출한다. 제안한 방법을 사용하여 도로 영상에서 실험한 결과를 보였다.

• 음성과학
• /
• 제12권3호
• /
• pp.105-114
• /
• 2005

It was verified that a speaker verification system improved its performances of EER by regularizing log likelihood ratio, using background speaker models. Recently the wireless mobile phones are becoming more dominant communication terminals than wired phones. So the need for building a speaker verification system on mobile phone is increasing abruptly. Therefore in this paper, we had some experiments to examine the performance of speaker verification based on mobile phone's voices. Especially we are focused on the performance variations in EER(Equal Error Rate) according to several background speaker's characteristics, such as selecting methods(MSC, MIX), number of background speakers, aging factor of speech database. For this, we constructed a speaker verification system that uses GMM(Gaussin Mixture Model) and found that the MIX method is generally superior to another method by about 1.0% EER. In aspect of number of background speakers, EER is decreasing in proportion to the background speakers populations. As the number is increasing as 6, 10 and 16, the EERs are recorded as 13.0%, 12.2%, and 11.6%. An unexpected results are happened in aging effects of the speech database on the performance. EERs are measured as 4%, 12% and 19% for each seasonally recorded databases from session 1 to session 3, respectively, where duration gap between sessions is set by 3 months. Although seasons speech database has 10 speakers and 10 sentences per each, which gives less statistical confidence to results, we confirmed that enrolled speaker models in speaker verification system should be regularly updated using the ongoing claimant's utterances.

• 대한전자공학회논문지SP
• /
• 제48권1호
• /
• pp.157-165
• /
• 2011

본 논문에서는 마르코프 랜덤 필드(Markov random fields: MRF) 기반으로 배경을 모델링하는 방식과 함께 관련 파라미터들을 추정하는 알고리즘을 제안한다. 화소 기반의 배경 모델링 기법은 인근 화소 간의 연관성을 고려하지 않고 화소 단위의 시간적 변화에 대한 통계적 특성에 주로 의존하므로 판정 오류를 줄이는데 한계가 있다. 제안 알고리즘은 화소 기반으로 배경 모델을 일차적으로 수행한 다음 MRF를 이용하여 시공간적으로 인근한 화소 간의 상호 의존성을 활용하여 배경모텔의 정확도를 향상시키는데 그 목적을 두고 있다. MRF는 기본적으로 파라미터의 크기에 매우 민감하므로 기존의 MRF 기반 알고리즘은 이미지에 따라 적절한 값을 사전에 구하여 적용하고 있다. 제안한 방식은 초기에 임의의 파라미터로 배경/전경 상태변수를 구한 후에 이의 통계적 특성을 이용하여 파라미터들을 추정하고 추정된 파라미터를 적용하여 상대변수를 재차 구하는 과정을 반복함으로써 최적의 파라미터에 적응적으로 수렴하도록 조정한다. 실내외의 다양한 환경에서 촬영한 비디오를 이용하여 제안한 방식 성능을 확인한다.

• 대한전자공학회:학술대회논문집
• /
• 대한전자공학회 2003년도 컴퓨터소사이어티 추계학술대회논문집
• /
• pp.111-114
• /
• 2003

In this paper, we propose a depth-based mesh modeling method to generate virtual environment. The proposed algorithm constructs mesh model from unorganized point cloud obtained from a multi-view camera. We separate the point cloud consisting objects from the background. Then, we apply triangulation to each object and background. Since the objects and the background are modeled independently, it is possible to construct effective virtual environment. The application of proposed modeling method is applicable to entertainment, such as movie and video game and effective virtual environment generation.

• Communications for Statistical Applications and Methods
• /
• 제8권1호
• /
• pp.299-309
• /
• 2001

In PET(Positron Emission Tomography), it is necessary to use transmission scan data in order to estimate the attenuation map. Recently, there are several empirical studies in which one might be able to estimate attenuation map and activity distribution simultaneously with emissive sinogram alone without transmission scan. However, their algorithms are based on the model in which does not include the background counts term, and so is unrealistic. If the background counts component has been included in the model, their algorithm would introduce non-monotonic reconstruction algorithm which results in vain in practice. in this paper, we develop a monotonic and parallelizable algorithm for simultaneous reconstruction of both characteristics and present the validity through some simulations.

• 대한의용생체공학회:의공학회지
• /
• 제8권2호
• /
• pp.223-230
• /
• 1987

In this paper the harmonic degrees between $\alpha$ wave and $\beta$ wave in visual evoked potential are analyzed by the bicoherence. The bicoherence analysis is based on an AR model which provides significantly better resolution than that of Fourier transform. The analysis results of visual evoked pope ntial are compared with the analysis results of background EEC. From the comparison results it is found that the harmonic degree of visual evoked potential is less than she harmonic degree of background EEG and the $\beta$ wave of visual evoke potential unlike the background EEC contains the non harmonic property of a wave more than the harmonic properity

• 한국지구물리탐사학회:학술대회논문집
• /
• 한국지구물리탐사학회 2007년도 공동학술대회 논문집
• /
• pp.197-202
• /
• 2007

Most of seismic reflection prospecting assumes subsurface formation to be homogeneous media. These models are not capable of estimating small scale heterogeneity which is verified by well log data or drilling core. And those synthetic seismograms by homogeneous media are limited to explain various changes at field data. So we developed a inhomogeneous velocity model which can estimate inhomogeneity of background medium to implement numerical modeling from homogeneous medium and inhomogeneous medium on the model. Background medium using three autocorrelation functions in order to generate inhomogeneous velocity media was according to dominant wavelength of background medium and correlation length of random medium. And then we compared shot gathers. The results show that numerical modeling implemented at inhomogeneous medium depicts complex wave propagation of field data.

• 스마트미디어저널
• /
• 제5권1호
• /
• pp.78-87
• /
• 2016

The major problems of recent object tracking methods are related to the inefficient detection of moving objects due to occlusions, noisy background and inconsistent body motion. This paper presents a robust method for the detection and tracking of a moving in infrared animal videos. The tracking system is based on adaptive optical flow generation, Gaussian mixture and Kalman filtering. The adaptive Gaussian model of optical flow (GMOF) is used to extract foreground and noises are removed based on the object motion. Kalman filter enables the prediction of the object position in the presence of partial occlusions, and changes the size of the animal detected automatically along the image sequence. The presented method is evaluated in various environments of unstable background because of winds, and illuminations changes. The results show that our approach is more robust to background noises and performs better than previous methods.