• Communications for Statistical Applications and Methods
• /
• 제6권3호
• /
• pp.831-842
• /
• 1999

We propose a wavelet decomposition feature extraction method for the hand-written character recognition. Comparing the recognition rates of which methods with original image features and with selected features by the wavelet decomposition we study the characteristics of the proposed method. LDA(Linear Discriminant Analysis) QDA(Quadratic Discriminant Analysis) RDA(Regularized Discriminant Analysis) and NN(Neural network) are used for the calculation of recognition rates. 6000 hand-written numerals from CENPARMI at Concordia University are used for the experiment. We found that the set of significantly selected wavelet decomposed features generates higher recognition rate than the original image features.

• 융합신호처리학회논문지
• /
• 제3권3호
• /
• pp.27-35
• /
• 2002

지식기반 영상검색은 영상이 갖는 다양한 데이터에서 추출되어진 특징값을 지식으로 하여 질의 영상에 대한 검색 결과영상을 찾아주는 방법이다. 본 연구에서 사용한 영상자료는 자동차 전조등 영상으로 전조등 영상에 대한 입력 자료는 차량마다 다양한 패턴을 갖는 영상과 문자, 숫자 및 특수문자이다. 영상에서의 정보는 화소값들의 분포상태나 통계적 분석 및 패턴의 상태 등인데, 전조등 영상에서는 이러한 정보가 영상 검색을 위한 지식 데이터로 사용된다. 영상데이터에서 추출된 다양한 정보를 다중 지식 기반으로 하여 본 논문에서는 교통사고나 기타 차량사건의 발생 시 활용할 수 있는 영상검색 시스템을 구축하였으며, 전조등 영상의 검색에 효율적으로 적용한 다중 지식기반 검색방법을 제안하였다. 다중지식 구축을 위한 특징함수는 컬러 영상에서와 그레이레벨 영상에서 각각 필요한 성분들을 추출하여 구성하였으며, 한 개나 두 개 정도의 특징값을 사용한 기존의 방법과 달리 복합적인 특징값의 사용을 통한 다중 지식 기반의 검색방법이 컬러정보나 패턴에 대한 유사성을 높여서 용의차량의 전조등 영상 검색 효율성을 향상시켰다. 소프트웨어의 제작을 위해 비쥬얼 베이직과 크리스탈리포트 그리고 MS 액세스 데이터베이스를 사용하였다. 검색 효율성 및 특성 함수의 구성을 효과적으로 발전시키면 검색시스템은 용의 차량의 추적 및 교통사고에서 효율적인 과학수사에 일조할 것으로 기대한다.

• 전자공학회논문지A
• /
• 제33A권5호
• /
• pp.203-211
• /
• 1996

This paper describes CAD tools for the construction of image database in IC chip analysis CAD system. For IC chip analysis by high-resolution microscopy, the image database is essential to manage more than several thousand images. But manual database construction is error-prone and time-consuming. In order to solve this problem, we develop a set of CAD toos that include image grabber to capture chip images, image editor to make the whole chip image database from the grabbed images, and image divider to reconstruct the database that consists of evenly overlapped images for efficient region search. we also develop an interactive pattern matching method for user-friendly image editing, and a heuristic region search method for fast image division. The tools are developed with a high-performance graphic hardware with JPEG image comparession chip to process the huge color image data. The tools are under the field test and experimental resutls show that the database construction time can be redcued in 1/3 compared to manual database construction.

• 한국의류산업학회지
• /
• 제4권1호
• /
• pp.11-18
• /
• 2002

The purpose of this study was to investigate the characteristic and composition factor of Japanese costume design. The stimulus were 25 contemporary costume design which represented the traditional image of Japanese. The main survey of questionary consisted of their evaluation of the Japanese costume image by 26 semantic differential bipolar scales and the subjects were 99 female students majoring in clothing and textiles. The data were analyzed by Factor analysis, Multidimensional Scaling Method and Regression Analysis. The major findings were as follows. As a result of design analysis, contemporary Japanese costume design which represented the traditional image had traditional form, color, texture, pattern, etc. Through factor analysis about Japanese costume image 7 factors were identified; Attractiveness, Attention, Cool and warm, Neatness, Activeness, Maturity, Classics. According to image positioning, Japanese costume design was classified by simple-decorative, soft-hard. As the result of regression analysis, The preference of Japanese costume image was related to attractive factor.

• 대한산업공학회지
• /
• 제41권3호
• /
• pp.267-274
• /
• 2015

In the semiconductor manufacturing post-FAB process, predicting a package test result accurately in the wafer testing phase is a key element to ensure the competitiveness of companies. The prediction of package test can reduce unnecessary inspection time and expense. However, an analysing method is not sufficient to analyze data collected at wafer testing phase. Therefore, many companies have been using a summary information such as a mean, weighted sum and variance, and the summarized data reduces a prediction accuracy. In the paper, we propose an analysis method for Wafer Map Image collected at wafer testing process and conduct an experiment using real data.

• 한국측량학회지
• /
• 제40권2호
• /
• pp.109-118
• /
• 2022

싱크홀 탐사 등 지반조사를 위한 기술로 개발된 GPR (Ground Penetrating Radar)은 지하시설물 탐사에서 불탐구간을 해소하기 위한 방법으로 한정되어 사용하고 있었다. 정부는 지하시설물 데이터의 정확도 개선을 위하여 2022년 7월부터 비금속 관로 탐사기를 이용한 지하시설물 탐사가 가능하도록 하였다. 그러나 GPR은 점토층 등과 같이 연약지반 같은 수분함량이 높은 지반에서 탐사율도 낮아지고, 정확도에 많은 변동이 발생하는 문제점을 가지고 있다. 본 연구에서는 GPR의 특성과 지하시설물의 환경을 고려한 탐사정확도 향상방안으로 유전률 상수 보정과 GPR 영상자료의 패턴분석을 이용한 지하시설물 GPR탐사 방안을 제시하고자 한다. 본 연구를 통하여 GPR 주파수 대역과 이기종 GPR을 적용한 현장검증 결과 지하시설물 탐사의 정확도 향상 및 높은 재현성 결과를 도출하였다.

• Nuclear Engineering and Technology
• /
• 제54권10호
• /
• pp.3943-3948
• /
• 2022

Single-photon emission computed tomography is one of the reliable pin-by-pin verification techniques for spent-fuel assemblies. One of the challenges with this technique is to increase the total fuel assembly verification speed while maintaining high verification accuracy. The aim of the present study, therefore, was to develop an artificial intelligence (AI) algorithm-based tomographic image analysis technique for partial-defect verification of fuel assemblies. With the Monte Carlo (MC) simulation technique, a tomographic image dataset consisting of 511 fuel-rod patterns of a 3 × 3 fuel assembly was generated, and with these images, the VGG16, GoogLeNet, and ResNet models were trained. According to an evaluation of these models for different training dataset sizes, the ResNet model showed 100% pattern estimation accuracy. And, based on the different tomographic image qualities, all of the models showed almost 100% pattern estimation accuracy, even for low-quality images with unrecognizable fuel patterns. This study verified that an AI model can be effectively employed for accurate and fast partial-defect verification of fuel assemblies.

• 전자공학회논문지CI
• /
• 제45권5호
• /
• pp.56-62
• /
• 2008

의사결정나무를 구성하여 강판튜브 비파괴평가에 사용하는 산업용 CR영상의 측정 패턴인식을 도모한다. 본래 비파괴평가는 기계학습기법에 의한 패턴식별과 그 분류에 적합한 분야이다. 의사결정나무의 속성들은 비파괴평가 테스트 절차로부터 취한다. 방사선조사 입사각, 경사도 및 거리 둥의 기하학적 특성들은 입력 영상 데이터 분석으로부터 추정한다. 이 요소들은 대상 입력을 의사결정나무에서 미리 정해진 분류에로 정확히 그리고 쉽게 분류가 이루어지도록 한다. 이 알고리즘은 비파괴평가 결과의 특성화를 간단히 하며 특성 결정을 간편하게 한다. 실험 결과는 제안한 알고리즘의 유용성을 보였다.

• 복식문화연구
• /
• 제22권3호
• /
• pp.399-410
• /
• 2014

The purpose of this study was to investigate emotional images and preference of knitwear by tone on tone combination. The subjects were 357 university students in Daejeon and Chungnam province, and the measuring instruments were 6 stimuli manipulated by color and tone combination type of background and pattern in the tone and tone combination, and self-administrated questionnaires consisted of emotional images items, preference items, and subjects' demographics attributions. The data were analyzed by Cronbach's ${\alpha}$, factor analysis, t-test, MANOVA and Duncan's multiple range test, using SPSS program. The results were as follows. First, four factors (attractiveness, conspicuity, mildness, and activity) are emerged on emotional images of knitwear. Second, color had main effects on emotional images and preference. Gray color was perceived as most attractive image and more preferred than others. Third, tone combination type had some effects on emotional images. Vivid tone background/light tone pattern was perceived more attractive image but less conspicuous and mild than light tone background/vivid tone pattern. Forth, subjects' gender had an effects on conspicuous image. Male was perceived more conspicuous image on knitwear stimuli than female. Fifth, color and subjects' gender had interaction effects on attractiveness image and preference. Male perceived that blue is more attractive and preferred than female.

• 한국ITS학회 논문지
• /
• 제13권2호
• /
• pp.27-33
• /
• 2014

This paper proposes a gradation image processing method for the development of a Road Sign Recognition Platform (RReP), which aims to facilitate the rapid and accurate management and surveying of approximately 160,000 road signs installed along the highways, national roadways, and local roads in the cities, districts (gun), and provinces (do) of Korea. RReP is based on GPS(Global Positioning System), IMU(Inertial Measurement Unit), INS(Inertial Navigation System), DMI(Distance Measurement Instrument), and lasers, and uses an imagery information collection/classification module to allow the automatic recognition of signs, the collection of shapes, pole locations, and sign-type data, and the creation of road sign registers, by extracting basic data related to the shape and sign content, and automated database design. Image division and merging, which were applied in this study, produce superior results compared with local binarization method in terms of speed. At the results, larger texts area were found in images, the accuracy of text recognition was improved when images had been gradated. Multi-threshold values of natural scene images are used to improve the extraction rate of texts and figures based on pattern recognition.