• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.18 no.2
• /
• pp.420-437
• /
• 2024

There are some problems in network traffic classification (NTC), such as complicated statistical features and insufficient training samples, which may cause poor classification effect. A NTC architecture based on one-dimensional Convolutional Neural Network (CNN) and transfer learning is proposed to tackle these problems and improve the fine-grained classification performance. The key points of the proposed architecture include: (1) Model classification--by extracting normalized rate feature set from original data, plus existing statistical features to optimize the CNN NTC model. (2) To apply transfer learning in the classification to improve NTC performance. We collect two typical network flows data from Youku and YouTube, and verify the proposed method through extensive experiments. The results show that compared with existing methods, our method could improve the classification accuracy by around 3-5%for Youku, and by about 7 to 27% for YouTube.

• Journal of the Korean Society of Surveying, Geodesy, Photogrammetry and Cartography
• /
• v.21 no.1
• /
• pp.19-26
• /
• 2003

As recently, high-resolution satellite images of 1m spatial resolution are opened to the public and able to be used commercially, the studies that make ortho-images using them and apply to digital mapping and database of geo-spatial information system are having been progressed actively. Therefore, the purposes of this study are to establish the auto-extraction methods and to develope algorithms for automatically extracting buildings out of man-made structures, after making the IKONOS ortho-image. As the result of this study, we can extract buildings automatically at 72% out of the whole buildings. And we have analyzed the error trend by means of the comparison with ortho-image, digital map and drawing result, then we could know that obtain the good result for extraction of the building through the methods and algorithms of this study.

• KSCE Journal of Civil and Environmental Engineering Research
• /
• v.14 no.1
• /
• pp.161-170
• /
• 1994

Traditional methods for the extraction of the environmental factors of waters in which environmental change is severer than in the land can not examine closely the changed phenomena because of the lack of equipments, manpower, time and cost, etc. Therefore, new practical and efficient methods are required. The research for the method to manage environment of the waters with remote sensing technology was needed. This study examined the interrelations between the data by an on-the-spot survey and Landsat TM data and presented the model for extracting factors of water quality with regression analysis and experimental formula.

• The Research Journal of the Costume Culture
• /
• v.28 no.4
• /
• pp.492-507
• /
• 2020

This study aims to develop fashion designs that combine atlas fabric with the characteristics of Uygur costume to modernize the costume of the Xinjiang Uygur. Research contents and methods are as follows. First, based on previous studies, research analysis was conducted on the cultural background, clothing characteristics, and material of Uygur clothing. Second, based on such research contents, designs combining the characteristics of Uygur costume and atlas fabric were presented. Third, to analyze the utilization of atlas fabric and examine fabric characteristics, material was gathered from collections on domestic and foreign web sites. Through field explorations of local museums in the Xinjiang area, minority group culture was observed in more detail. Based on collection of traditional clothing and analysis of its characteristics, fashion designs that apply contemporary trends were developed. General silhouettes without any restrictions to the waist and decorations made using embroidery were often used. Atlas silk, developed in China using Ikat weaving methods, is an important traditional clothing fabric of the minority group Xinjiang. Based on such data collection analysis, the produced works highlighted traditional ethnic characteristics by extracting classical patterns of atlas fabric, modifying or partially expanding them, combining them with hand knitting, and adding contemporary sensations, thus providing confirmations of the possibility of popularizing classic patterns in more practical manners.

• International Journal of Fuzzy Logic and Intelligent Systems
• /
• v.16 no.4
• /
• pp.246-253
• /
• 2016

Infrared (IR) imaging has been researched for various applications such as surveillance. IR radiation has the capability to detect thermal characteristics of objects under low-light conditions. However, automatic segmentation for finding the object of interest would be challenging since the IR detector often provides the low spatial and contrast resolution image without color and texture information. Another hindrance is that the image can be degraded by noise and clutters. This paper proposes multi-level segmentation for extracting regions of interest (ROIs) and objects of interest (OOIs) in the IR scene. Each level of the multi-level segmentation is composed of a k-means clustering algorithm, an expectation-maximization (EM) algorithm, and a decision process. The k-means clustering initializes the parameters of the Gaussian mixture model (GMM), and the EM algorithm estimates those parameters iteratively. During the multi-level segmentation, the area extracted at one level becomes the input to the next level segmentation. Thus, the segmentation is consecutively performed narrowing the area to be processed. The foreground objects are individually extracted from the final ROI windows. In the experiments, the effectiveness of the proposed method is demonstrated using several IR images, in which human subjects are captured at a long distance. The average probability of error is shown to be lower than that obtained from other conventional methods such as Gonzalez, Otsu, k-means, and EM methods.

• Journal of Distribution Science
• /
• v.14 no.2
• /
• pp.5-12
• /
• 2016

Purpose - The present study examines the interrelationship between various components constituting shopping experience in the context of the Indian shopping malls. Research design, data, and methodology - Extracting components of shopping experience from the literature review, the study used Interpretive Structural Modelling (ISM) to propose a conceptual model. The study adopted a mixed methods research involving theoretical constructs from past research, qualitative assessment of relationship between the constructs and imposing definite order and direction to qualitative relations based on mathematical computations. Results - Proposed model indicates that the five components of shopping experience (ambience, physical infrastructure, convenience, marketing focus and safety and security) do not converge directly into shopping experience. Rather, they operate following a hierarchy of influences in which marketing focus plays the role of the initiator. Conclusions - This model points at the order of preference of different components of shopping experience and can be a useful guide for retail industry, especially mall developers and supermarket/hypermarket, may use the findings in key decisions about development of physical infrastructure, which are based on marketing focus.

• Journal of Sensor Science and Technology
• /
• v.25 no.4
• /
• pp.285-290
• /
• 2016

Human sensing is an important but challenging technology. Unlike other methods for sensing humans, a vision sensor has many advantages, and there has been active research in automatic human detection in camera images. The combination of Histogram of Oriented Gradients (HOG) and Support Vector Machine (SVM) is currently one of the most successful methods in vision-based human detection. However, extracting HOG features from an image is computer intensive, and it is thus hard to employ the HOG method in real-time processing applications. This paper describes an efficient solution to this speed problem of the HOG method. Our method obtains edge information of an image and finds candidate regions where humans very likely exist based on the distribution pattern of the detected edge points. The HOG features are then extracted only from the candidate image regions. Since complex HOG processing is adaptively done by the guidance of the simpler edge detection step, human detection can be performed quickly. Experimental results show that the proposed method is effective in various images.

• Journal of Information Technology Services
• /
• v.13 no.4
• /
• pp.397-410
• /
• 2014

For mobile service organizations, it is one of the most important tasks to reflect users' opinions rapidly and accurately. In this study, the process is defined to elicit requirements of software/system improvement for mobile application by extracting and refining from users' opinion in mobile app, and detailed activities procession method in this processing are also proposed. The process consists of 3 activities to get requirements of software/system improvement for mobile app. First activity is to transform mobile app to software structure and define term dictionary. Second activity is to elicit simple sentences based on software from users' opinion and refine them. The last activity is to integrate and adjust refined requirements. To verify the usability and validity of the proposed process and the methods, the outputs of manual processing and semi-automated processing were compared. As a result, efficiency and improvement possibility of the process were confirmed through extraction ratio of requirements, comparison of execution time, and analysis of agreement ratio.

• Transactions of the Korean Society for Noise and Vibration Engineering
• /
• v.22 no.8
• /
• pp.742-747
• /
• 2012

A modified NDIF method using a sub-domain approach is introduced to extract highly accurate eigenvalues of two-dimensional, arbitrarily shaped acoustic cavities. The NDIF method, which was developed by the authors for the eigen-mode analysis of arbitrarily shaped acoustic cavities, has the feature that it yields highly accurate eigenvalues compared with other analytical methods or numerical methods(FEM and BEM). However, the NDIF method has the weak point that it can be applicable for only convex cavities. It was revealed that the solution of the NDIF method is very inaccurate or is not suitable for concave cavities. To overcome the weak point, the paper proposes the sub-domain method of dividing a concave domain into several convex domains. Finally, the validity of the proposed method is verified in two case studies, which indicate that eigenvalues obtained by the proposed method are more accurate compared to the exact method, the NDIF method, or FEM(ANSYS).

• Proceedings of the Korean Society of Precision Engineering Conference
• /
• 2002.05a
• /
• pp.684-687
• /
• 2002

The authors have developed two devices which cuts the spend fuel rod to an optimal size and extracts fuel pellet from the pieces of cut fuel rods. These devices are so important to reduce radioactive wastes that some advanced countries developed their own methods and devices. The authors have benchmarked from these methods and devices. For spent fuel rod cutting, the tube cutting method has been chosen. some mechanical properties of the fuel tube and pellet has been carefully considered for an optimal cutting size. For fuel pellet extraction, a mechanically extracting method has been adopted. The existing chemical method have turned out to be inappropriate because it produced large amount of radioactive wastes, in spite of its high fuel recovery characteristics. The developed method has an advantage that it can be applied to other fuel rods that have different shapes and sizes. The two devices are set up and operated in the hot cell where people can not go in, so that the devices have been designed to be controlled remotely and modulated for easy maintenance. And the performance of the devices has been tested by using simulated fuel rod. From the experimental results, the devices are supposed to be useful for reducing radioactive wastes.