• Smart Structures and Systems
• /
• 제31권4호
• /
• pp.409-419
• /
• 2023

Structural integrity can be accessed from dynamic deformations of structures. Moreover, dynamic deformations can be acquired from non-contact sensors such as video cameras. Kanade-Lucas-Tomasi (KLT) algorithm is one of the commonly used methods for motion tracking. However, averaging throughout the extracted features would induce bias in the measurement. In addition, pixel-wise measurements can be converted to physical units through camera intrinsic. Still, the depth information is unreachable without prior knowledge of the space information. The assigned homogeneous coordinates would then mismatch manually selected feature points, resulting in measurement errors during coordinate transformation. In this study, a two-stage optimization method for video-based measurements is proposed. The manually selected feature points are first optimized by minimizing the errors compared with the homogeneous coordinate. Then, the optimized points are utilized for the KLT algorithm to extract displacements through inverse projection. Two additional criteria are employed to eliminate outliers from KLT, resulting in more reliable displacement responses. The second-stage optimization subsequently fine-tunes the geometry of the selected coordinates. The optimization process also considers the number of interpolation points at different depths of an image to reduce the effect of out-of-plane motions. As a result, the proposed method is numerically investigated by using a truss bridge as a physics-based graphic model (PBGM) to extract high-accuracy displacements from recorded videos under various capturing angles and structural conditions.

• International Journal of Korean Welding Society
• /
• 제1권2호
• /
• pp.23-29
• /
• 2001

In this study, a preview-sensing visual sensor system is constructed far weld seam tracking in GMA welding. The visual sensor system consists of a CCD camera, a diode laser system with a cylindrical lens, and a band-pass-filter to overcome the degrading of image due to spatters and/or arc light. Among the image processing methods, Hough transform method is compared with the central difference method from a viewpoint of the capability for extracting the accurate feature position. As a result, it was revealed that Hough transform method can more accurately extract the feature positions and it can be applied to real time weld seam tracking. Image processing which includes Hough transform method is carried out to extract straight lines that express laser stripe. After extracting the lines, weld joint position and edge points are determined by intersecting the lines. Even though the image includes a spatter trace on it, it is possible to recognize the position of weld joint. Weld seam tracking was precisely implemented with adopting Hough transform method, and it is possible to track the weld seam in the case of offset angle is in the region of $\pm$ $15^{\circ}$.

• Journal of Information Processing Systems
• /
• 제19권4호
• /
• pp.483-497
• /
• 2023

Depth estimation is one of the most complicated and difficult problems to deal with in the light field. In this paper, a compound attention convolutional neural network (CAttNet) is proposed to extract depth maps from light field images. To make more effective use of the sub-aperture images (SAIs) of light field and reduce the redundancy in SAIs, we use a compound attention mechanism to weigh the channel and space of the feature map after extracting the primary features, so it can more efficiently select the required view and the important area within the view. We modified various layers of feature extraction to make it more efficient and useful to extract features without adding parameters. By exploring the characteristics of light field, we increased the network depth and optimized the network structure to reduce the adverse impact of this change. CAttNet can efficiently utilize different SAIs correlations and features to generate a high-quality light field depth map. The experimental results show that CAttNet has advantages in both accuracy and time.

• Journal of Korea Artificial Intelligence Association
• /
• 제1권2호
• /
• pp.1-6
• /
• 2023

In this paper, we present a new method for classifying malicious URLs to reduce cases of learning difficulties due to unfamiliar and difficult terms related to information protection. This study plans to extract only visually distinguishable features within the URL structure and compare them through map learning algorithms, and to compare the contribution values of the best map learning algorithm methods to extract features that have the most impact on classifying malicious URLs. As research data, Kaggle used data that classified 7,046 malicious URLs and 7.046 normal URLs. As a result of the study, among the three supervised learning algorithms used (Decision Tree, Support Vector Machine, and Logistic Regression), the Decision Tree algorithm showed the best performance with 83% accuracy, 83.1% F1-score and 83.6% Recall values. It was confirmed that the contribution value of https is the highest among whether to use https, sub domain, and prefix and suffix, which can be visually distinguished through the feature contribution of Decision Tree. Although it has been difficult to learn unfamiliar and difficult terms so far, this study will be able to provide an intuitive judgment method without explanation of the terms and prove its usefulness in the field of malicious URL detection.

• 정보처리학회논문지B
• /
• 제13B권7호
• /
• pp.689-698
• /
• 2006

패턴인식은 주위 환경을 관찰하는 방법, 배경으로부터 관심있는 패턴을 구분하는 방법, 소리를 얻는 방법, 그리고 패턴 범주들 중에서 타당한 결정을 얻는 방법에 관한 연구이다. 패턴인식 시스템을 설계할 때 필수적으로 1) 데이터의 획득과 전처리, 2) 데이터의 표현, 3) 결정방법 선택과 같은 세 가지 사항을 고려해야한다. 그 이유는 영상을 획득하기 위한 센서의 선택, 전처리 기법, 표현 기법, 의사결정 모델에 따라 인식의 결과가 달라질 수 있기 때문이다. 컬러영상은 다양한 컬러 패턴으로 구성된다. 대부분의 패턴인식 방법은 훈련되어진 컬러정보를 사용하여 컬러의 특징을 추출한다. 본 논문은 몇 가지 제한된 컬러를 가진 영상으로부터 특정한 컬러 패턴을 적응적으로 추출한다. 컬러 패턴의 수가 한정되어 있기 때문에 영상에서 컬러의 분포가 유사하다. 그러나, 영상에 잡음이나 열화가 존재하면, 그 분포가 변화한다. 그러므로 이미 알고 있는 컬러정보를 가지고 특정한 컬러의 특징을 추출할 수 없다. 그래서 본 논문에서는 유사한 컬러 패턴을 가진 영상에 대하여 특정한 컬러의 특징을 적응적으로 추출함으로서 인식의 오류를 감소시킬 수 있는 새로운 방법을 제안한다. 제안한 방법을 실험하기 위하여 열화가 적은 표본영상을 사용하고, 잡음과 열화가 포함된 여섯 가지의 검사영상을 사용한다. 결론적으로 제안한 방법이 통계적인 패턴인식의 결과보다 정확한 결과를 보여준다.

• Journal of Biosystems Engineering
• /
• 제24권2호
• /
• pp.159-166
• /
• 1999

A system to extract continuously the real 3-D geometric fearture information from 2-D image of an object, which is fed randomly via conveyor has been developed. Two sets of structured laser lightings were utilized. And the laser structured light projection image was acquired using the camera from the signal of the photo-sensor mounted on the conveyor. Camera coordinate calibration matrix was obtained, which transforms 2-D image coordinate information into 3-D world space coordinate using known 6 points. The maximum error after calibration showed 1.5 mm within the height range of 103mm. The correlation equation between the shift amount of the laser light and the height was generated. Height information estimated after correlation showed the maximum error of 0.4mm within the height range of 103mm. An interactive 3-D geometric feature extracting software was developed using Microsoft Visual C++ 4.0 under Windows system environment. Extracted 3-D geometric feature information was reconstructed into 3-D surface using MATLAB.

• 대한전자공학회논문지SP
• /
• 제40권3호
• /
• pp.217-226
• /
• 2003

본 논문에서는 세선화 지문 영상의 순차적 레이블링을 이용하여 위치 이동, 크기 변화 그리고 회전에 무관한 새로운 지문 융선 특징 검출 알고리즘을 제안한다. 제안한 알고리즘은 먼저 지문의 중심점을 지나는 수직선을 이용하여 세선화 지문 영상의 융선을 순차적으로 레이블링 한다. 그리고 레이블링한 개개의 융선들로부터 특징을 검출한다 검출하는 특징은 융선의 종류와 융선에 존재하는 특징점의 융선 각도이다. 이러한 방법을 이용하여 지문 융선의 특징을 검출하면, 지문을 이루고 있는 여러 융선들의 종류를 알 수 있고, 각 융선에 존재하는 특징점의 종류 및 이들의 각도를 알 수 있다. 두 개의 세선화 지문 영상을 이용하여 실험한 결과, 제안하는 알고리즘이 위치 이동, 크기 변화 그리고 회전에 무관한 지문 융선 특징을 검출함을 확인하였다.

• 한국정보처리학회논문지
• /
• 제3권4호
• /
• pp.899-907
• /
• 1996

본 연구에서는 교차수를 이용한 시스템을 향상하기 위하여 오류 특징점을 제거하기 위한 알고리즘을 제안하였다. 제안된 알고리즘을 이용하므로써 특징점을 정확하게 추출할 수 있었고 인식율도 증가할 수 있었다.제안 알고리즘의 검증을 위한 실험 대상은 잡음을 부분적, 그리고 전체적으로 추가한 화상과, 절단한 화상등 3그룹을 이용하였다. 그 결과 부분적 잡음이나, 전체적 잡음이 있고, 절단된 화상일 경우 제안한 알고리 즘을 적용하였을 때 각각 97.7%, 97.7%, 95% 의 인식율을 보였다. 따라서 제안한 알 고리즘을 적용한 후 인식하는게 타당하리라 사료된다.

• 한국컴퓨터그래픽스학회논문지
• /
• 제12권4호
• /
• pp.17-22
• /
• 2006

본 논문은 주어진 얼굴 사진에 대하여 자동으로 얼굴의 특징을 강조한 캐리커쳐 생성 기법을 제안한다. AAM(Active Appearance Model)을 사용하여 트레이닝 이미지의 특징점과 텍스쳐 정보를 유지하고 이것을 이용하여 평균 얼굴의 정보와 함께 주어진 얼굴에 대한 특징점을 찾아낸다. 캐리커쳐 아티스트들의 제안을 바탕으로 특징적인 부분을 과장하기 위한 룰을 정의하고 이를 입력 얼굴의 특징점에 적용하여 과장된 특징점을 얻는다. 마지막으로 주어진 사진에 대하여 좀 더 만화적인 효과를 내기 위해 얼굴 이미지에 카투닝을 적용한 다음 과장된 특징점으로 와핑한다. 이러한 방법으로 사용자의 조작을 최소로 하는 캐리커처 생성을 할 수 있다.

• Journal of Electrical Engineering and Technology
• /
• 제13권6호
• /
• pp.2485-2492
• /
• 2018

The rapid development of large-scale integrated circuits has brought great challenges to the circuit testing and diagnosis, and due to the lack of exact fault models, inaccurate analog components tolerance, and some nonlinear factors, the analog circuit fault diagnosis is still regarded as an extremely difficult problem. To cope with the problem that it's difficult to extract fault features effectively from masses of original data of the nonlinear continuous analog circuit output signal, a novel approach of feature extraction and dimension reduction for analog circuit fault diagnosis based on wavelet packet decomposition, local linear embedding algorithm, and clone selection algorithm (WPD-LLE-CSA) is proposed. The proposed method can identify faulty components in complicated analog circuits with a high accuracy above 99%. Compared with the existing feature extraction methods, the proposed method can significantly reduce the quantity of features with less time spent under the premise of maintaining a high level of diagnosing rate, and also the ratio of dimensionality reduction was discussed. Several groups of experiments are conducted to demonstrate the efficiency of the proposed method.