• Journal of the Computational Structural Engineering Institute of Korea
• /
• v.36 no.5
• /
• pp.339-346
• /
• 2023

The uncertainties of microstructural features affect the properties of materials. Numerous pores that are randomly distributed in materials make it difficult to predict the properties of the materials. The distribution of pores in cementitious materials has a great influence on their mechanical properties. Existing studies focus on analyzing the statistical relationship between pore distribution and material responses, and the correlation between them is not yet fully determined. In this study, the mechanical response of cementitious materials is predicted through an image-based data approach using a convolutional neural network (CNN), and the correlation between pore distribution and material response is analyzed. The dataset for machine learning consists of high-resolution micro-CT images and the properties (tensile strength) of cementitious materials. The microstructures are characterized, and the mechanical properties are evaluated through 2D direct tension simulations using the phase-field fracture model. The attributes of input images are analyzed to identify the spot with the greatest influence on the prediction of material response through CNN. The correlation between pore distribution characteristics and material response is analyzed by comparing the active regions during the CNN process and the pore distribution.

• KIPS Transactions on Software and Data Engineering
• /
• v.7 no.1
• /
• pp.9-18
• /
• 2018

Recently, the use of image stitching technology has been increasing as the number of contents based on virtual reality increases. Image Stitching is a method for matching multiple images to produce a high resolution image and a wide field of view image. The image stitching is used in various fields beyond the limitation of images generated from one camera. Image Stitching detects feature points and corresponding points to match multiple images, and calculates the homography among images using the RANSAC algorithm. Generally, corresponding points are needed for calculating conversion relation. However, the corresponding points include various types of noise that can be caused by false assumptions or errors about the conversion relationship. This noise is an obstacle to accurately predict the conversion relation. Therefore, RANSAC algorithm is used to construct an accurate conversion relationship from the outliers that interfere with the prediction of the model parameters because matching methods can usually occur incorrect correspondence points. In this paper, we propose an algorithm that extracts more accurate inliers and computes accurate transformation relations by using correspondence point relation information used in RANSAC algorithm. The correspondence point relation information uses distance ratio between corresponding points used in image matching. This paper aims to reduce the processing time while maintaining the same performance as RANSAC.

• Journal of the Korea Society of Computer and Information
• /
• v.26 no.4
• /
• pp.21-28
• /
• 2021

This study proposes a deep learning-based flat foot classification methodology using transfer learning. We used a transfer learning with VGG16 pre-trained model and a data augmentation technique to generate a model with high predictive accuracy from a total of 176 image data consisting of 88 flat feet and 88 normal feet. To evaluate the performance of the proposed model, we performed an experiment comparing the prediction accuracy of the basic CNN-based model and the prediction model derived through the proposed methodology. In the case of the basic CNN model, the training accuracy was 77.27%, the validation accuracy was 61.36%, and the test accuracy was 59.09%. Meanwhile, in the case of our proposed model, the training accuracy was 94.32%, the validation accuracy was 86.36%, and the test accuracy was 84.09%, indicating that the accuracy of our model was significantly higher than that of the basic CNN model.

• Journal of the Korea Computer Graphics Society
• /
• v.29 no.1
• /
• pp.13-22
• /
• 2023

In this paper, we present a method for estimating full 3D shapes from given 2D silhouette images of human bodies. Because the silhouette only consists of the partial information on the true shape, it is an ill-posed problem. To address the problem, we use the statistical human shape space obtained from the existing large 3D human shape database. The method consists of three steps. First, we extract the boundary pixels and their appropriate normal vectors from the input silhouette images. Then, we initialize the correspondences of each pixel to the vertex of the statistically-deformable 3D human model. Finally, we numerically optimize the parameters of the statistical model to fit best to the given silhouettes. The viability and the robustness of the method is demonstrated with various experiments.

• Journal of the Korea Convergence Society
• /
• v.11 no.10
• /
• pp.207-212
• /
• 2020

This study analyzes the Soju(Korean alcohol) advertisement image, which is relatively easy to interpret subjectively, among sexual appeal advertisements that stimulate consumers' curiosity, where the image is verified through AOI (area of interest) 3 areas (face, body, product), and eye-tracking, one of the psychophysiological indicators. The result of the analysis reveals that visual attention, the interest in the advertising model, was higher in the face than in the body shape. Contrary to the prediction that men would be more interested in body shape than women, both men and women showed higher interest in the face than a body. Besides, it was derived that recognition and recollection of the product were not significant. This study is significant in terms of examining the pattern of visual attention such as the gaze point and gaze time of male and female consumers on sexual appeal advertisements. In further, the study looks forward to bringing a positive influence to the soju advertisement image by presenting the expression method that the soju advertisement image should pursue as well as the appropriate marketing direction.

• Journal of Internet of Things and Convergence
• /
• v.9 no.1
• /
• pp.65-70
• /
• 2023

In this paper, we propose an autonomous driving system using an end-to-end model to improve lane departure and misrecognition of traffic lights in a vision sensor-based system. End-to-end learning can be extended to a variety of environmental conditions. Driving data is collected using a model car based on a vision sensor. Using the collected data, it is composed of existing data and data with outlayers removed. A class was formed with camera image data as input data and speed and steering data as output data, and data learning was performed using an end-to-end model. The reliability of the trained model was verified. Apply the learned end-to-end model to the model car to predict the steering angle with image data. As a result of the learning of the model car, it can be seen that the model with the outlayer removed is improved than the existing model.

• Annual Conference of KIPS
• /
• 2023.05a
• /
• pp.590-591
• /
• 2023

최근 사람들의 건강에 대한 관심이 많아지고 다양한 운동 컨텐츠가 확산되면서 실내에서 운동을 할 수 있는 기회가 많아졌다. 하지만, 전문가의 도움없이 정확하지 않은 동작을 수행하다 큰 부상을 입을 위험성이 높다. 본 연구는 CNN 기반 요가 자세 분류 모델을 생성하고 설명가능 인공지능 기술을 적용하여 예측 결과에 대한 해석을 제시한다. 사용자에게 설명성과 신뢰성 있는 모델을 제공하여 자신에게 맞게 올바른 자세를 결정할 수 있고, 무리한 동작으로 부상을 입을 확률 또한 낮출 수 있을 것으로 보인다.

• Proceedings of the Korean Society of Computer Information Conference
• /
• 2023.07a
• /
• pp.403-404
• /
• 2023

옷감의 종류에 따른 올바른 세탁 방법이 존재하는데, 이를 따르지 않고 세탁을 하게 되면 옷이 금방 손상된다. 이러한 잘못된 세탁 방법으로 세탁하여 옷이 손상되는 문제점을 해소하고자 세탁물 인식 및 세탁 방법제공 웹사이트를 제작하였다. 개발된 웹사이트는 사전에 학습된 모델을 바탕으로 사용자의 세탁물 이미지를 인식하여 예측 결과에 따른 세탁 방법을 제공하며 이를 통해 사용자는 성분이 불분명한 옷감에 대한 정보와 세탁 방법을 얻을 수 있다.

• Proceedings of the Korea Water Resources Association Conference
• /
• 2023.05a
• /
• pp.507-507
• /
• 2023

지진해일은 많은 인명피해를 입힐 수 있는 위험한 자연재해이며, 예를 들어 각각 약 25만명과 약 2만명의 사상자가 발생하였던 2004년 수마트라 지진해일과 2011년 동일본 지진해일 등이 있다. 우리나라 동해안 또한 향후 지진 발생 가능성이 큰 지진공백역이 존재하여 안전한 지역으로 볼 수 없다. 지진해일 방재대책 수립과 관련된 연구는 지속적으로 이루어지고 있지만 지진해일의 발생빈도는 적고 완벽히 대응하는 것은 현실적으로 불가능하다. 따라서 본 연구에서는 지진해일 방재대책의 가장 기본적인 자료로 이용될 수 있는 지진해일 침수예상도를 효율적인 방법으로 제작하는 것을 연구했다. 현재 우리나라의 지진해일 최대 침수예상도는 과거 및 향후 발생가능한 지진해일의 경우에 대한 모든 범람구역이 고려된 보수적인 방법으로 제작되고 있다. 지진원의 위치와 각 매개변수의 특성에 따라 범람구역이 다양하게 나타날 수 있기 때문에 보수적인 최대 침수예상도는 과도한 범람구역이 고려될 수 있다. 따라서 본 연구에서는 보수적인 최대 침수예상도와 비교하여 AI기술과 로직트리 기법을 통해 더 정확한 최대 침수예상도를 제작하는 것을 목표로 한다. 연구방법은 1) 고려된 모든 지진해일 시나리오에 대한 수치해석 2) 입력자료인 지진해일 초기수면 변위 이미지 증강 3) CNN모델을 활용한 초기수면 변위 이미지 분류 4) 분류된 결과의 범람 구역으로 최대 침수예상도를 제작하였다. 향후 연구결과는 지진해일 재해정보도 제작 및 지진해일 침수예측 모델 개발에 활용될 수 있을 것이다.

• Spatial Information Research
• /
• v.22 no.2
• /
• pp.63-71
• /
• 2014

Ocean leisure sports have recently emerged as one of so-called blue ocean industries. They are sensitive to diverse environmental conditions such as current, temperature, and salinity, which can increase needs of forecasting data as well as in-situ observations for the ocean. In this context, a Web-based geovisualization system for coastal information produced by model forecasts was implemented for use in supporting various ocean activities. First, FVCOM(Finite Volume Coastal Ocean Model) was selected as a forecasting model, and its data was preprocessed by a spatial interpolation and sampling library. The interpolated raster data for water surface elevation, temperature, and salinity were stored in image files, and the vector data for currents including speed and direction were imported into a distributed DBMS(Database Management System). Web services in REST(Representational State Transfer) API(Application Programming Interface) were composed using Spring Framework and integrated with desktop and mobile applications developed on the basis of hybrid structure, which can realize a cross-platform environment for geovisualization.