• 제어로봇시스템학회논문지
• /
• 제16권7호
• /
• pp.695-702
• /
• 2010

This paper presents an image alignment algorithm for application of AOI (Automatic Optical Inspection) based on SIFT. Since the correspondences result using SIFT descriptor have many wrong points for aligning, this paper modified and classified those points by five measures called the CCFMR (Cascade Classifier for False Matching Reduction) After reduced the false matching, rotation and translation are estimated by point selection method. Experimental results show that the proposed method has fewer fail matching in comparison to commercial software MIL 8.0, and specially, less than twice with the well-controlled environment’s data sets (such as AOI system). The rotation and translation accuracy is robust than MIL in the noise data sets, but the errors are higher than in a rotation variation data sets although that also meaningful result in the practical system. In addition to, the computational time consumed by the proposed method is four times shorter than that by MIL which increases linearly according to noise.

• 한국환경보건학회지
• /
• 제28권1호
• /
• pp.31-39
• /
• 2002

Generally, inaccurate reports on environmental issues occur due to various factors. The purpose of this study was to find out a way to enhance accuracy of environmental reporting. So the reporters' career and college major had been compared to the accuracy of their articles. The by-lined environmental articles in nine dailies published in 1991 were checked. Results of this study were as follows. (I) Inaccuracy rate in environmental articles was 54.2%. Inaccuracies appeared 1.7 times per an article, while the average frequency of inaccuracies in overall articles was 0.9 time. (2) Errors in the articles consist of 65.8% of subjective inaccuracies and 34.2% of subjective inaccuracies. They derive from the false usage of terminology(15.8% ),misquotation(14.5%), incorrect statistics(13%), exaggeration(13%), inaccurate title(7.9%), and false comparison(5%). (3) Inaccuracy rate by the type of articles was 66.7% in columns, 60% in feature stories, 54.5% in-depth stories, 40.9% in straight news, respectively. (4) Inaccuracy rate by the specific field was shown 71.4% in environmental impacts assessment, 52.5% in water pollution, 37.5% in waste management, and 35.7% in air pollution. (5) According to the result of chi-square test analysis, there were no statistically significant differences of inaccuracy rate and of subjective, and objective inaccuracies relevant to the period of reporters' career covering environmental reporting and the nature of articles, and college major.

• 한국수학교육학회지시리즈A:수학교육
• /
• 제49권2호
• /
• pp.161-174
• /
• 2010

The purpose of the study is to analyze various types of errors appeared in true-false proof problems of matrix and to find out correction method. In order to achieve this purpose, error test was conducted to the subject of 87 second grade students who were chosen from D high schoool. It was shown from this test that the most frequent error type was caused by the lack of understanding about concepts and essential facts of matrix(35.3%), and then caused by the invalid logically reasoning (27.4%), and then caused by the misusing conditions(18.7%). Through three hours of correction lessons with 5 students, the following correction teaching method was proposed. First, it is stressed that the operation rules and properties satisfied in real number system can not be applied in matrix. Second, it is taught that the analytical proof method and the reductio ad absurdum method are useful in the proof problem of matrix. Third, it is explained that the counter example of E=$\begin{pmatrix}1\;0\\0\;1 \end{pmatrix}$, -E should be found in proof of the false statement. Fourth, it is taught that the determinant condition should be checked for the existence of the inverse matrix.

• International Journal of Advanced Culture Technology
• /
• 제9권4호
• /
• pp.260-267
• /
• 2021

In this paper, we intend to use a new mixed model of YoloV5 and DeepSort. For fire detection, we want to increase the accuracy by automatically extracting the characteristics of the flame in the image from the training data and using it. In addition, the high false alarm rate, which is a problem of fire detection, is to be solved by using this new mixed model. To confirm the results of this paper, we tested indoors and outdoors, respectively. Looking at the indoor test results, the accuracy of YoloV5 was 75% at 253Frame and 77% at 527Frame, and the YoloV5+DeepSort model showed the same accuracy at 75% at 253 frames and 77% at 527 frames. However, it was confirmed that the smoke and fire detection errors that appeared in YoloV5 disappeared. In addition, as a result of outdoor testing, the YoloV5 model had an accuracy of 75% in detecting fire, but an error in detecting a human face as smoke appeared. However, as a result of applying the YoloV5+DeepSort model, it appeared the same as YoloV5 with an accuracy of 75%, but it was confirmed that the false positive phenomenon disappeared.

• 대한안전경영과학회지
• /
• 제25권4호
• /
• pp.87-94
• /
• 2023

This study analyzed and identified various causes of caustic alarms of 163 fire detectors that occurred from January 2019 to December 2021 at domestic semiconductor manufacturing plants equipped with about 30,000 fire detectors, and proposed a new non-fire prevention cause investigation plan by applying the NFPA 921 scientific methodology. The results of the study are as follows. First, in terms of necessary recognition and problem definition, an analog detector and an integrated monitoring system were proposed to quickly determine the location and installation space information of the fire detector. Second, in order to prevent speculative causes and errors in various analyses in terms of data analysis and hypothesis establishment, non-fire reports were classified into five by factor and defined, and the causes of occurrence by factor were classified and proposed. Finally, in terms of hypothesis verification and final hypothesis selection, a non-fire prevention improvement termination process and a final hypothesis verification sheet were proposed to prevent the cause from causing re-error.

• 한국통신학회논문지
• /
• 제30권11B호
• /
• pp.730-735
• /
• 2005

현재 인터넷 상에서 신속한 의사소통을 위해 사용되는 전자우편의 증가는 상업적 의도를 가진 상품 홍보 수단으로 악용되며, 많은 사회적 문제를 유발시키고 있다. 이에 다양한 스팸 차단 필터 기술이 개발되고 있으나, 차단필터의 성능에 따라 정상 메일을 스팸 메일로 오인하여 사용자의 시스템 이용 신뢰도를 크게 저하시키고 있다. 본 논문에서는 스팸 메일 차단 시스템의 이용 신뢰도를 높이기 위해 Privacy 기반의 스팸 메일 복구 기법이 적용된 SMBC(Spam Mail Blocking Center) 스팸 메일 차단 플랫폼을 설계 및 제안한다. SMBC는 Proxy Server 기반의 스팸 차단 시스템 프레임 레이어로 설계되며, 물리적으로 임의의 위상(Topology)로 구축 가능하여, 플랫품 구현시 유연한 모듈/구성 레이어 개발이 가능하다. 제안된 SMBC 플랫폼은 기존 스팸 메일 차단 시스템에 비해 처리 부하와 차단 필터의 오인률을 최소화하여 시스템 이용 신뢰도를 높일 수 있도록 설계되었다.

• 한국정보과학회논문지:소프트웨어및응용
• /
• 제32권5호
• /
• pp.357-365
• /
• 2005

본 연구에서는 유사한 여러 물체들이 인접하여 나타나는 영상열로부터 물체들을 개별적으로 분리할 수 있는 B-spline 적합(fitting) 알고리즘을 제안한다. 기존의 스네이크(snake) 알고리즘들은 초기화의 어려움과 다수의 극점 존재로 인해서 이러한 영상자료에서 물체의 영역을 개별적으로 분리하는 데는 어려움이 있다. 본 연구에서는 이 문제를 극복하고 다양한 형태의 물체가 인접해 있는 유사한 물체들로부터 효과적으로 분할 할 수 있는 유전자(genetic) 알고리즘 기반 B-spline 적합방안을 제안한다. 실제 상황을 고려하여 생성된 영상자료와 실제 치아 CT 영상을 이용한 평가에서 제안된 방법은 서로 인접해 있는 유사한 형태와 자기의 물체들을 개별적으로 정확하게 분할할 수 있음을 보였다. 제안된 알고리즘의 결과는 이상적으로 추출된 영역과의 일치성과 false positive 오류 그리고 false negative오류가 계산되어 검증되었다.

• 정보보호학회논문지
• /
• 제25권5호
• /
• pp.1043-1057
• /
• 2015

지능형 위협을 빠르게 인지하고 능동적으로 탐지 및 대응하기 위해 주요 공공단체 및 민간기관에서는 침입탐지시스템(IDS)을 관리 운영하고 있으며, 이는 공격의 검출 및 탐지에 매우 중요한 역할을 한다. 그러나 IDS 경보의 대부분은 오탐(false positive)을 생성하는 문제가 있다. 또한, 알려지지 않은 악성코드를 탐지하고 사전에 위협을 인지 대응하기 위해서 APT대응솔루션이나 행위기반체계를 도입 운영하고 있다. 이는 가상기술을 이용해 악성코드를 직접실행하고 가상환경에서 이상행위를 탐지하거나 또는 다른방식으로 알려지지 않은 공격을 탐지한다. 그러나 이 또한 가상환경 회피, 트래픽 전수조사에 대한 성능적 문제, 정책오류 등의 약점 등이 존재한다. 이에 따라 결과적으로 효과적인 침입탐지를 위해서는 보안관제 고도화가 매우 중요하다. 본 논문에서는 보안관제 고도화의 한가지 방안으로 침입탐지시스템의 주요 단점인 오탐(false positive)을 줄이는 방안에 대해 논한다. G기관의 경험적 데이터를 근거로 실험을 수행한 결과 세 가지 유형 11가지 규칙을 도출하였다. 이 규칙을 준수하여 테스트한 결과 전반적인 오탐율이 30%~50% 이상 줄어들고 성능이 30% 이상 향상됨을 검증하였다.

• 제어로봇시스템학회:학술대회논문집
• /
• 제어로봇시스템학회 2003년도 ICCAS
• /
• pp.2594-2599
• /
• 2003

We consider some problems of the Modified SPRT(Sequential Probability Ratio Test) method for fault detection and isolation of inertial redundant sensor systems and propose an Advanced SPRT method to solve the problems of the Modified SPRT method. One problem of the Modified SPRT method to apply to inertial sensor system comes from the effect of inertial sensor errors such as misalignment, scale factor error and sensor bias in the parity vector, which make the Modified SPRT method hard to be applicable. The other problem is due to the correlation of parity vector components which may induce false alarm. We use a two-stage Kalman filter to remove effects of the inertial sensor errors and propose the modified parity vector and the controlled parity vector which removes the effect of correlation of parity vector components. The Advanced SPRT method is derived form the modified parity vector and the controlled parity vector. Some simulation results are presented to show the usefulness of the Advanced SPRT method to redundant inertial sensor systems.

• 제어로봇시스템학회논문지
• /
• 제22권9호
• /
• pp.686-692
• /
• 2016

A Crimp Force Monitor (CFM) is equipment for detecting crimp errors by analyzing crimp signals obtained from force and strain sensors. The analysis is commonly performed by aligning a measured crimp signal with a reference signal and comparing their difference. Current analysis methods often suffer from wrong alignments that result in false negative detections. This paper presents a new crimp signal analysis method in CFM. First, a falling edge alignment is proposed that matches falling edges of the measured and the reference signals by minimizing the absolute difference summation. Second, a signal parameter error is introduced to evaluate the crimp quality difference between the measured signal and the reference. For calculating the signal parameter error, part of a signal is identified and divided into several regions to maximize the signal parameter errors. Experiments showed that the proposed method can improve the signal alignment and accurately detect bad crimps especially with the strain sensor.