• 전기학회논문지P
• /
• 제52권1호
• /
• pp.35-41
• /
• 2003

In this paper, a machine vision system for recognizing and classifying the patterns and marks engraved by die molding or laser marking on the glass panels of computer monitors is suggested and evaluated experimentally. The vision system is equipped with a neural network and an NGC pattern classifier including searching process based on normalized grayscale correlation and adaptive binarization. This system is found to be applicable even to the cases in which the segmentation of the pattern area from the background using ordinary blob coloring technique is quite difficult. The inspection process is accomplished by the use of the NGC hypothesis and ANN verification. The proposed pattern recognition system is composed of three parts: NGC matching process and the preprocessing unit for acquiring the best quality of binary image data, a neural network-based recognition algorithm, and the learning algorithm for the neural network. Another contribution of this paper is the method of generating the training patterns from only a few typical product samples in place of real images of all types of good products.

• 한국경영과학회지
• /
• 제30권4호
• /
• pp.27-43
• /
• 2005

This paper presents a new hybrid data mining technique using error pattern modeling to improve classification accuracy when the data type of a target variable is binary. The proposed method increases prediction accuracy by combining two different supervised learning methods. That is, the algorithm extracts a subset of training cases that are predicted inconsistently by both methods, and models error patterns from the cases. Based on the error pattern model, the Predictions of two different methods are merged to generate final prediction. The proposed method has been tested using practical 10 data sets. The analysis results show that the performance of proposed method is superior to the existing methods such as artificial neural networks and decision tree induction.

• Journal of Electrical Engineering and Technology
• /
• 제9권1호
• /
• pp.344-351
• /
• 2014

In order to accomplish complex navigational commands, humanoid robot should be able to modify its walking period, step length and direction independently. In this paper, a novel walking pattern generation algorithm is proposed to satisfy these requirements. Modification of the walking pattern can be considered as a transition between two periodic walking patterns, which follows each navigational command. By assuming the robot as a linear inverted pendulum, the equations of motion between ZMP(Zero Moment Point) and CM(Center of Mass) state is easily derived and analyzed. After navigational command is translated into the desired CM state, corresponding CM motion is generated to achieve the desired state by using simple ZMP functions. Moreover, when the command is not feasible, feasible command is alternated by using binary search algorithm. Subsequently, corresponding CM motion is generated. The effectiveness of the proposed algorithm is verified by computer simulation.

• 한국정밀공학회:학술대회논문집
• /
• 한국정밀공학회 2002년도 추계학술대회 논문집
• /
• pp.345-350
• /
• 2002

In this study, researchers classifying the artificial flaws in semiconductor packages are performed by pattern recognition technology. For this purposes, image pattern recognition package including the user made software was developed and total procedure including ultrasonic image acquisition, equalization filtration, binary process, edge detection and classifier design is treated by Backpropagation Neural Network. Specially, it is compared with various weights of Backpropagation Neural Network and it is compared with threshold level of edge detection in preprocessing method fur entrance into Multi-Layer Perceptron(Backpropagation Neural network). Also, the pattern recognition techniques is applied to the classification problem of defects in semiconductor packages as normal, crack, delamination. According to this results, it is possible to acquire the recognition rate of 100% for Backpropagation Neural Network.

• 한국정밀공학회:학술대회논문집
• /
• 한국정밀공학회 2005년도 추계학술대회 논문집
• /
• pp.721-726
• /
• 2005

In this study, researchers classifying the artificial flaws in semiconductor. packages are performed by pattern recognition technology. For this purposes, image pattern recognition package including the user made software was developed and total procedure including ultrasonic image acquisition, equalization filtration, binary process, edge detection and classifier design is treated by Backpropagation Neural Network. Specially, it is compared with various weights of Backpropagation Neural Network and it is compared with threshold level of edge detection in preprocessing method for entrance into Multi-Layer Perceptron(Backpropagation Neural network). Also, the pattern recognition techniques is applied to the classification problem of defects in semiconductor packages as normal, crack, delamination. According to this results, it is possible to acquire the recognition rate of 100% for Backpropagation Neural Network.

• 한국멀티미디어학회논문지
• /
• 제18권9호
• /
• pp.1008-1018
• /
• 2015

In this paper, we propose a method that constructs a saliency map in which important regions are accurately specified and the colors of the regions are less influenced by the similar surrounding colors. Our method utilizes LBP(Local Binary Pattern) histogram information to compare and analyze texture information of surrounding regions in order to reduce the effect of color information. We extract the saliency of stereoscopic images by integrating a 2D saliency map with depth information of stereoscopic images. We then measure the distance between two different sizes of the LBP histograms that are generated from pixels. The distance we measure is texture difference between the surrounding regions. We then assign a saliency value according to the distance in LBP histogram. To evaluate our experimental results, we measure the F-measure compared to ground-truth by thresholding a saliency map at 0.8. The average F-Measure is 0.65 and our experimental results show improved performance in comparison with existing other saliency map extraction methods.

• 한국정보통신학회논문지
• /
• 제17권9호
• /
• pp.2145-2152
• /
• 2013

제지 제조 산업은 대규모 설비가 요구되는 장치산업으로서 생산 설비의 자동화가 꼭 요구된다. 특히 제조공정의 효율성을 얻기 위해서는 제지 제조 공정 중에서 발생하는 지절의 결함을 효과적으로 검출하고 이를 분류하는 효율적인 요소 기술을 필요로 한다. 본 논문에서는 기존의 제지 제조 공정 방식의 문제점을 제시하고, 이를 효과적으로 개선하기 위하여 국부 이진 패턴 분석에 의한 지절 결함 검출 시스템을 제안하고 구현된 결과를 제시한다. 제안한 시스템은 제지 지절 결함에 대해 국부 이진 패턴 분석법을 이용하여 분류하고 이를 인식하는 방식으로 구성된다. 제안된 시스템은 에지형과 영역형 결함으로 지절 결함으로 분류하고, 현장 시스템에 설치되어 안정적인 결과를 보임이 검증되었다.

• Journal of Computing Science and Engineering
• /
• 제10권3호
• /
• pp.85-94
• /
• 2016

We propose a model-based method for information alignment using educational standards as a case study. Discrepancies and inconsistencies in educational standards across different states/cities hinder the retrieval and sharing of educational resources. Unlike existing educational standards alignment systems that only give binary judgments (either "aligned" or "not-aligned"), our proposed system classifies each pair of educational standard statements in one of seven levels of alignments: Strongly Fully-aligned, Weakly Fully-aligned, Partially-$aligned^{***}$, Partially-$aligned^{**}$, Partially-$aligned^*$, Poorly-aligned, and Not-aligned. Such a 7-level categorization extends the notion of binary alignment and provides a finer-grained system for comparing educational standards that can broaden categories of resource discovery and retrieval. This study continues our previous use of mathematics education as a domain, because of its generally unambiguous concepts. We adopt a materialization pattern (MP) model developed in our earlier work to represent each standard statement as a verb-phrase graph and a noun-phrase graph; we align a pair of statements using graph matching based on Bloom's Taxonomy, WordNet, and taxonomy of mathematics concepts. Our experiments on data sets of mathematics educational standards show that our proposed system can provide alignment results with a high degree of agreement with domain expert's judgments.

• 디지털융복합연구
• /
• 제11권4호
• /
• pp.259-266
• /
• 2013

패턴인식 등에서 널리 이용되는 이진연결요소 라벨링은 오래전부터 연구되어온 영상처리분야의 기본적인 처리방법이다. 연결요소 라벨링에 대한 현재의 연구는 이중스캔을 이용한 방법이 주류를 이루고 있는데 최근 일차스캔시 인근 화소 여러 개를 한 번에 블록단위로 처리하는 것이 가장 성능이 뛰어난 것으로 보고되고 있다. 본 논문에서도 블록단위의 라벨링 방법을 이용하였지만 기존의 방법들보다 더 성능이 개선된 방법을 제시하고 있다. 제안된 방법은 블록단위의 라벨과 새로운 화소기반의 스캔마스크를 사용했는데 실험결과 현재까지 발표된 가장 빠른 라벨링 방법보다도 더 우수한 성능을 보이는 것으로 나타났다.

• 한국통신학회논문지
• /
• 제39A권2호
• /
• pp.92-99
• /
• 2014

본 논문은 결정 트리(Decision tree) 구조를 기반으로 한 표정 인식 방법을 제안한다. ASM(Active Shape Model)과 LBP(Local Binary Pattern)를 통해, 표정 영상들의 국소 특징들을 추출한다. 국소 특징들로부터 표정들을 잘 분류할 수 있는 판별 특징(Discriminant feature)들을 추출하고, 그 판별 특징들은 모든 조합의 각 두 가지 표정들을 분류시킨다. 분류를 통해 얻어진 정인식의 합을 통해, 정인식 최대화 기반 국소 영역과 표정 조합을 결정한다. 이 가지 분류들을 종합하여, 결정 트리를 생성한다. 이 결정 트리 기반 표정 인식률은 약 84.7%로, 결정 트리를 고려하지 않은 방법보다, 더 좋은 인식 성능을 보였다.