• KSTLE International Journal
• /
• v.1 no.1
• /
• pp.34-42
• /
• 2000

A method and results of classification of four different metallic wear debris were presented by using their color features. The color image of wear debris was used far the initial data, and the color properties of the debris were specified by HSI color model. Particles were characterized by a set of statistical features derived from the distribution of HSI color model components. The initial feature set was optimized by a principal component analysis, and multidimensional scaling procedure was used fer the definition of a classification plane. It was found that five features, which include mean values of H and S, median S, skewness of distribution of S and I, allow to distinguish copper based alloys, red and dark iron oxides and steel particles. In this work, a method of probabilistic decision-making of class label assignment was proposed, which was based on the analysis of debris-coordinates distribution in the classification plane. The obtained results demonstrated a good availability for the automated wear particle analysis.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.8 no.12
• /
• pp.4489-4501
• /
• 2014

We propose a novel lane detection method based on classification in image sequences. Both structural and statistical features of the extracted bright shape are applied to the neural network for finding correct lane marks. The features used in this paper are shown to have strong discriminating power to locate correct traffic lanes. The traffic lanes detected in the current frame is also used to estimate the traffic lane if the lane detection fails in the next frame. The proposed method is fast enough to apply for real-time systems; the average processing time is less than 2msec. Also the scheme of the local illumination compensation allows robust lane detection at nighttime. Therefore, this method can be widely used in intelligence transportation systems such as driver assistance, lane change assistance, lane departure warning and autonomous vehicles.

• Korean Journal of Bronchoesophagology
• /
• v.9 no.2
• /
• pp.49-55
• /
• 2003

Background and Objectives : Traditionally incision and drainage is considered to be standard treatment of deep neck infection. But antibiotics and diagnostic technique are developed recently, there are reports that conservative therapy could be as successful as open surgical drainage. The purposes of this study has been to assess clinical feature between surgical therapy group and conservative therapy group through statistical analysis. Materials and Methods : A retrospective study was performed on 46 cases of deep neck space abscess, which were confirmed CT, in patients admitted from January 1999 to June 2002. Result : About 80% of all are treated with conservative therapy. Erythrocyte sediment rate, volume of abscess and duration of hospitalization of conservative therapy group are decreased than those of surgical therapy group. Conclusions : Conservative therapy is expected to be effective on treatment of early stage, small sized deep neck infection. But its complication can lead to serious condition of patient it should be done under meticulous observation.

• Annual Conference on Human and Language Technology
• /
• 1992.10a
• /
• pp.237-248
• /
• 1992

본 논문에서는 통계적인 특징 추출 방법을 사용하여 샘물체로 필기된 한글을 고속으로 인식하는 방법을 소개한다. 대부분이 직선 성분으로 이루어진 한글의 특성을 이응하기 위하여 입력 영상으로부터 수평, 수직, 사선, 역사선의 방향 성분을 추출하며, 검은 화소의 밀도에 따라 동적으로 그물을 결정함으로써 획 간의 접촉 변형에 무관한 특징 벡터를 추출한다. 이와 같은 통계적 특징 추출 방법은 크기 정규화나 세선화 과정이 필요없으며, 또한 샘물체라는 필기 형태의 제약에 의해 정합 대상 부류의 수가 현저히 줄어들기 때문에 인식에 소요되는 시간을 상당히 줄일 수 있음은 물론, 인식률을 향상시켰다. 제안된 방법의 타당성을 검증하기 위하여 샘물체로 필기된 KS 완성형 한글 2,350자에 대해 실험한 결과, 평균 90% 이상의 인식률을 보이며, IBM PC 486(33MHz)상에서 문자당 평균 0.17초의 인식 속도를 보임으로써, 실용적인 고속 OCR 시스템의 개발 가능성을 확인할 수 있었다.

• Proceedings of the Computational Structural Engineering Institute Conference
• /
• 2007.04a
• /
• pp.135-140
• /
• 2007

In this paper, an Improved Dimension Reduction(IDR) method is proposed for uncertainty quantification that employes Kriging interpolation technic. It has been acknowledged that the DR method is accurate and efficient for assessing statistical moments and reliability due to the sensitivity free feature. However, the DR method has a number of drawbacks such as instability and inaccuracy for problems with increased nonlineality. In this paper, improved DR is implanted by three steps. First, the Kriging interpolation method is used to accurately approximate the responses. Second, 2N+1 and 4N+1 ADOEs are proposed to maintain high accuracy of the method for UQ analysis. Third, numerical integration scheme is used with accurate but free response values at any set of integration points of the surrogated model.

• ETRI Journal
• /
• v.33 no.3
• /
• pp.454-457
• /
• 2011

In this letter, we propose a new quality measurement method to identify the causes of video quality degradation for IP-based video services. This degradation mainly results from network performance issues and video compression. The proposed algorithm identifies the causes based on statistical feature values from blocky block distribution in degraded IP-based videos. We found that the sensitivity and specificity of the proposed algorithm are 93.63% and 91.99%, respectively, in comparison with real error types and subjective test data.

• Proceedings of the IEEK Conference
• /
• summer
• /
• pp.259-262
• /
• 2004

In most of the character recognition systems, the method of template matching or statistical method using hidden Markov model is used to extract and recognize feature shapes. In this paper, we used modified chain-code which has 8-directions but 4-codes, and made the chain-code of hand-written character, after that, converted it into transition chain-code by applying to HMM(Hidden Markov Model). The transition chain code by HMM is analyzed as signal flow graph by Mason's theory which is generally used to calculate forward gain at automatic control system. If the specific forward gain and feedback gain is properly set, the forward gain of transition chain-code using Mason's theory can be distinguished depending on each object for recognition. This data of the gain is reorganized as tree structure, hence making it possible to distinguish different hand-written characters. With this method, $91\%$ recognition rate was acquired.

• Journal of Information Technology Applications and Management
• /
• v.10 no.4
• /
• pp.135-147
• /
• 2003

With the development of multimedia and optical technologies, application systems with facial features hare been increased the interests of researchers, recently. The previous research efforts in face processing mainly use the frontal images in order to recognize human face visually and to extract the facial expression. However, applications, such as image database systems which support queries based on the facial direction and image arrangement systems which place facial images automatically on digital albums, deal with the directional characteristics of a face. In this paper, we propose a method to detect facial directions by using facial features. In the proposed method, the facial trapezoid is defined by detecting points for eyes and a lower lip. Then, the facial direction formula, which calculates the right and left facial direction, is defined by the statistical data about the ratio of the right and left area in facial trapezoids. The proposed method can give an accurate estimate of horizontal rotation of a face within an error tolerance of $\pm1.31$ degree and takes an average execution time of 3.16 sec.

• Proceedings of the Korean Institute of Electrical and Electronic Material Engineers Conference
• /
• 1999.05a
• /
• pp.234-238
• /
• 1999

The correlation between shape of electrical trees and partial discharge(PD) pulses in low density polyethylene(LDPE) were discussed. We observed growth feature of electrical tree by using optical microscope. On the basis of experimental results of measurements of trees occurring in the needle-plane arrangement with needle shape void and without needle shape void , statistical quantities are derived, which are relevant to PD pulse amplitude and phase. The PD quantities detected by partial discharge detector. we were analyzed q-n distribution pattern and $\psi$ -q-n distribution pattern. In this experiment, electrical trees in the needle-plane arrangement with needle shape void propagated branch type tree and in the needle-plane arrangement without needle shape void propagated bush type tree

• Journal of the Korea Safety Management & Science
• /
• v.7 no.5
• /
• pp.59-67
• /
• 2005

The main objective of the statistical analysis about industrial accidents is to find out what is the dangerous factor in its own industrial field so that it is possible to prevent or decrease the number of the possible accidents by educating those who work in the fields for safety tools. However, so far, there is no technique of quantitative evaluation on danger. Almost all previous researches as to industrial accidents have only relied on the frequency analysis such as the analysis of the constituent ratio on accidents. As an application of data mining technique, this paper presents analysis on the efficiency of the CHAID algorithm to classify types of industrial accidents data and thereby identifies potential weak points in accident risk grouping.