• Journal of KIISE:Software and Applications
• /
• v.36 no.8
• /
• pp.673-682
• /
• 2009

Recently, photo sharing and publishing based Social Network Sites(SNSs) are increasingly attracting the attention of academic and industry researches. Unlike the face recognition environment addressed by existing works, face annotation problem under SNSs is differentiated in terms of daily updated images database, a limited number of training set and millions of users. Thus, conventional approach may not deal with these problems. In this paper, we proposed a face annotation method for sharing and publishing photographs that contain faces under a social network service using random projection, non-linear regression and representational state transfer. Our experiments on several databases show that the proposed method records an almost constant execution time with comparable accuracy of the PCA-SVM classifier.

• Journal of Families and Better Life
• /
• v.27 no.3
• /
• pp.201-212
• /
• 2009

The purpose of research is to investigate patterns and diffusion of consumer participation on Internet UCC based on digital contents services. A sample survey of internet users was conducted, responses were collected from 629 respondents and consumer streaming data were analyzed. Some of the practical implications of the results are follows. Research can find out that patterns of user participation in UCC. The major genres of UCC are like daily lives of individuals, humors, parodies of star entertainers and types of contents like still pictures or images, texts are relatively highly generated comparing with multimedia UCC. Although participants have been being increased in UCC recently, the consumers as prosumers who are classified in contents generating group are ten percents at the most. In generating community-based UCC such as posting answers of questions and activities in Blog, prosumers who are in contents making group(recreational group) show more positive attitudes than simple participants(consuming only). The results of multiple regression analysis indicated that fun & entertainment, arousal, self-expression, user friendly web interface variable commonly posited a significant effect in multimedia UCC Services between two groups. Information sharing and perceived usefulness posited a significant effect in recreational group.

• Journal of the Korea Society of Computer and Information
• /
• v.25 no.9
• /
• pp.31-36
• /
• 2020

This research study designed a location image collecting technology. It provides the exact location information of an image which is not given in the photo to the user. Deep learning technology analysis and collects the images. The purpose of this service system is to provide the exact place name, location and the various information of the place such as nearby recommended attractions when the user upload the image photo to the service system. Suggested system has a deep learning model that has a size of 25.3MB, and the model repeats the learning process 50 times with a total of 15,266 data, performing 93.75% of the final accuracy. This system can also be linked with various services potentially for further development.

• Journal of Korea Multimedia Society
• /
• v.3 no.3
• /
• pp.209-223
• /
• 2000

As the importance and utilization of multimedia data increases, it becomes necessary to represent and manage multimedia data within database systems. In this paper, we designed and implemented an image retrieval system which support efficient management and intelligent retrieval of image data using concept hierarchy and data mining techniques. We stored the image information intelligently in databases using concept hierarchy. To support intelligent retrievals and efficient web services, our system automatically extracts and stores the user information, the user's query information, and the feature data of images. The proposed system integrates user metadata and image metadata to support various retrieval methods on image data.

• Korean Institute of Interior Design Journal
• /
• v.24 no.1
• /
• pp.169-177
• /
• 2015

The study aimed to classify the types of domestic dormitory furniture, and analyze the characteristics of each furniture type based on the empirical research methodology. The study consisted of literature review and field visit followed by survey and in-depth interview. The researchers collected 140 images of furniture from the web sites of 87 universities. Using affinity diagram, the dormitory furniture was classified into four different types: single fixed type, single semi-fixed type, multi fixed type, and multi semi-fixed type. The finding showed that the use of single fixed type was dominant in the domestic dormitory room, which had competitive price and easy maintenance. Both single fixed type and multi fixed type turned out to be lack of storage space. Meanwhile, both single semi-fixed type and multi semi-fixed type got the high value on the space efficiency due to the multi function. However, these two types could only be applied to a wide space enough for the furniture to be transformed and extended. The study analyzed the main characteristics of the dormitory furniture according to the type classification, and it is expected that this empirical study could work as a medium and database for the upcoming dormitory furniture design studies.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2011.05a
• /
• pp.119-122
• /
• 2011

The Stereo vision using belief propagation algorithm that has been studied recently yields good performance in disparity extraction. In this paper, BP algorithm is proved theoretically to high precision for a stereo matching algorithm. We derive disparity map from stereo image by using Belief Propagation (BP) algorithm and area-based matching algorithm. Two algorithms are compared using stereo images provided by Middlebury web site. Disparity map error rate decreased from 52.3% to 2.3%.

• 한국정보디스플레이학회:학술대회논문집
• /
• 2005.07b
• /
• pp.1611-1614
• /
• 2005

In this paper, a new multi-view 3D video communication system for real-time Reality teleconferencing application is proposed by usin gthe IEEE 1394 digital cameras, Intel Xeon server computer system and Microsoft's DirectShow programming library and its performance is analyzed in terms of image-grabbing frame rate and number of views. The captured two-view image data is compressed by extraction of disparity data between them and transmitted to another client system through the communication network, in which multi-view could be synthesized with this received 2-view data using the intermediate view reconstruction technique and displayed on the multi-view 3D display system. From some experimental results, it is found that the proposed system can display 16-view 3D images with a gray of 8bits and a frame rate of 15fps.

• Interdisciplinary Bio Central
• /
• v.1 no.1
• /
• pp.5.1-5.4
• /
• 2009

This article is an addendum to the authors’ previous article (Kim, J. et al. (2008) Plant Cell 20, 307-319). The instrumentation and software described in this article were used to analyze the circadian leaf movement in the previous article. Here, we provide detailed and practical information on the instrumentation and the software. The source code of the LMA program is freely available from the authors. The circadian clock regulates a wide range of cyclic physiological responses with a 24 hour period in most organisms. Rhythmic leaf movement in plants is a typical robust manifestation of rhythms controlled by the circadian clock and has been used to monitor endogenous circadian clock activity. Here, we introduce a relatively easy, inexpensive, and simple approach for measuring leaf movement circadian rhythms using a USB-based web camera, public domain software and a Leaf Movement Assay (LMA) program. The LMA program is a semi-automated tool that enables the user to measure leaf lengths of individual Arabidopsis seedlings from a set of time-series images and generates a wave-form output for leaf rhythm. This is a useful and convenient tool for monitoring the status of a plant's circadian clock without an expensive commercial instrumentation and software.

• Fashion & Textile Research Journal
• /
• v.14 no.3
• /
• pp.363-370
• /
• 2012

This study is for the development of fashion cultural products that simultaneously evolved with the contemporary use of traditional temple culture in addition, it analyzed the cultural products available in the Korean market. Methodology, this study conducted a literature review and empirical research. We targeted the cultural products carried at twelve web-based shopping malls for Buddhist cultural products and six souvenir shops in Korean Buddhist temples to collect data on those products in order to analyze the items, design motives, materials, and price ranges. The study results showed that interior items represented the largest portion of the targeted goods, followed by accessories/sundries, clothing/fashion items, stationery, and tableware. The most commonly used design motive was lotuses, followed by the images of Buddha or Buddhist Goddesses and Dharma. The most common materials include fibers, jewelry (such as gold and silver), wood, metals, ceramics, paper, and plastic. The most active price range was between KRW10,000 and KRW50,000, followed by less than KRW10,000 and KRW100,000 to less than KRW500,000. This study discovered the potential for traditional temple culture to advance it further in a contemporary manner and indicated the need to develop a wide variety of cultural products and emphasize its global acceptance.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.14 no.3
• /
• pp.962-975
• /
• 2020

The accurate segmentation of infant brain MR image into white matter (WM), gray matter (GM), and cerebrospinal fluid (CSF) is very important for early studying of brain growing patterns and morphological changes in neurodevelopmental disorders. Because of inherent myelination and maturation process, the WM and GM of babies (between 6 and 9 months of age) exhibit similar intensity levels in both T1-weighted (T1w) and T2-weighted (T2w) MR images in the isointense phase, which makes brain tissue segmentation very difficult. We propose a deep network architecture based on U-Net, called Triple Residual Multiscale Fully Convolutional Network (TRMFCN), whose structure exists three gates of input and inserts two blocks: residual multiscale block and concatenate block. We solved some difficulties and completed the segmentation task with the model. Our model outperforms the U-Net and some cutting-edge deep networks based on U-Net in evaluation of WM, GM and CSF. The data set we used for training and testing comes from iSeg-2017 challenge (http://iseg2017.web.unc.edu).