• IEIE Transactions on Smart Processing and Computing
• /
• v.6 no.4
• /
• pp.262-268
• /
• 2017

Intelligent video surveillance systems have been developed to monitor global areas and find specific target objects using a large-scale database. However, person re-identification presents some challenges, such as pose change and occlusions. To solve the problems, this paper presents an improved person re-identification method using sparse representation and saliency-based dictionary construction. The proposed method consists of three parts: i) feature description based on salient colors and textures for dictionary elements, ii) orthogonal atom selection using cosine similarity to deal with pose and viewpoint change, and iii) measurement of reconstruction error to rank the gallery corresponding a probe object. The proposed method provides good performance, since robust descriptors used as a dictionary atom are generated by weighting some salient features, and dictionary atoms are selected by reducing excessive redundancy causing low accuracy. Therefore, the proposed method can be applied in a large scale-database surveillance system to search for a specific object.

• Proceedings of the Safety Management and Science Conference
• /
• 2007.11a
• /
• pp.27-37
• /
• 2007

Preliminary investigation of CC process indicates that the process is feasible and has significant potential in construction automation approach of large objects. The process aims at automated construction of whole houses as well as sub-components. The potential of CC became evident from the initial investigations and experiments with various materials and geometries. Using this process, a single house or a colony of houses, each with possibly a different design, may be automatically constructed in a single run as shown on

• The Journal of the Institute of Internet, Broadcasting and Communication
• /
• v.16 no.1
• /
• pp.33-42
• /
• 2016

Recently, an environment in which public opinions are expressed about various areas is expanded around SNSs or internet potals, thus, opinion documents get bigger rapidly. Under these circumstances, it is essential to utilize automatic summarization techniques for understanding whole contents of large opinion documents. However, it is hard to summarize efficiently those documents with traditional text summarization technologies since the documents include subject expressions as well as features of targets objects. Proposed method in this paper defines features of opinion documents, and designed to retrieve representative sentences expressing opinions of those features. In addition, through experiments, we prove the usefulness of proposed method.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2013.05a
• /
• pp.597-598
• /
• 2013

A lot of fire and crime accidents are caused to a significant national loss. For example, the network and power facilities in national industry facilities, the fire risk region in large scale factories such as nuclear and thermal power plants, large-sized buildings, cultural properties, metal and steel mills, chemical plants, oil refineries. The development of a fire detection system that can detects the temperature and movement of objects as high-level quality is essential to prevent these incidents and accidents fundamentally. In this paper, the fire detection system based on infrared thermal imaging & CCD camera id designed to solve these problems.

• Journal of Fashion Business
• /
• v.24 no.2
• /
• pp.44-59
• /
• 2020

This investigated the romanticism of Alessandro Michele who leads new trends in fashion industry as a creative director for Gucci. In this study, 307 women's wears exhibited by Alessandro Michele for 3 years in Milan collections from 2015 F/W to 2017 F/W are categorized according to the aesthetic characteristics of romanticism. Additionally, the five aesthetic features of Gucci's fashion design are an addition since the appearance of abnormally large size under the influence of postmodernism. The results based on the study purpose are as follows. First, the decoration makes the garments stand out by using fancy materials, splendid multicolor combination, applique embroidery and three-dimensional objects. Second, exotic style is expressed from an occidental perspective using Chinese traditional fashion or textiles, and exotic accessories. Thirdly, sensuality is manifested using see-through materials to emphasize women's breasts, or the beautiful and sexual expression of women's body. Fourth, the new creations are designed by combining contrasting elements, such as disparate materials and, different genders. Furthermore, traditional cultures like baroque and rococo, and the 1970's vintage looks inspires the creation of retro-style. Finally, the appearance of extremely large shoulders, silhouettes, and details of the human body can be explained with expansive tendency. The aforementioned results further suggest that diverse characteristics are expressed via reflection of contemporary art and trends.

• Journal of Institute of Control, Robotics and Systems
• /
• v.19 no.12
• /
• pp.1089-1093
• /
• 2013

Depth extraction using the structured illumination method is popularly applied since it has the benefit of measuring the object without contact. With multiple spatial frequencies and phase-shifting techniques, it is possible to extract the depth of objects with large discontinuity. For applications such as 3D (Three Dimensional) displays, 3D information of the object is required and is useful if corresponding to each view of the display. For this purpose, hypercentric optics is appropriate to measure the depth information of an object with a large field of view that is applicable for a 3D display. By experiment, we present the feasibility for phase-shifting profilometry using hypercentric optics to obtain the depth information of an object with the field of view appropriate for a 3D display.

• International Journal of Internet, Broadcasting and Communication
• /
• v.10 no.2
• /
• pp.7-12
• /
• 2018

The spread of IoT (Internet of Things) technology that connects objects based on wired / wireless networks is accelerating, and IoT-based smart home technology that constitutes a super connected network connecting sensors and home appliances existing inside and outside the home is getting popular. In addition, demand for alternative energy technologies such as photovoltaic power generation is rapidly increasing due to rapid increase of consumption of energy resources. Recently, small solar power systems for general households as well as large solar power systems for installation in large buildings are being introduced, but they are effectively implemented due to limitations of small solar panels and lack of power management technology. In this paper, we have studied smart home structure and IoT / IoL device discovery algorithm for energy harvesting system based on photovoltaic power generation, It is possible to construct an efficient smart home system for device control.

• The Bulletin of The Korean Astronomical Society
• /
• v.39 no.1
• /
• pp.62.2-62.2
• /
• 2014

We compiled near-infrared photometric and polarimetric catalog of sources in ~3 $9^{\prime}{\times}69^{\prime}$ size field in the eastern side of the Large Magellanic Cloud (LMC). This catalog lists all 1969 sources which are brighter than 14 mag and which signal to noise ratio of degree of polarization is greater than 3 in the J, H, and Ks bands. The photometric and polarimetric data were simultaneously obtained in J, H, and Ks bands using SIRPOL, an imaging polarimeter of the InfraRed Survey Facility (IRSF), in 2008 December and 2011 December. We classified the objects into several groups based on their locations on the color-magnitude diagram and compared their general properties of polarization. We measured wavelength dependence of this field to verify interstellar polarization of the LMC which occurred from the dichroic extinction. We also discuss the polarization structure in this field with the results from molecular cloud studies.

• The Bulletin of The Korean Astronomical Society
• /
• v.38 no.1
• /
• pp.53-53
• /
• 2013

We present preliminary results of far-infrared(FIR) study of the supernova remnant(SNR)s in the Large Magellanic Cloud using the Herschel HERITAGE (HERschel Inventory of The Agents of Galaxy Evolution) data set. HERITAGE provides FIR data covering the entire LMC at 100,160, 250, 350, and 500 um. In order to confirm FIR emission associated with SNRs, we refer to Magellanic Cloud Emission-Line Survey (MCELS) H-alpha & SII data, Spitzer surveying the Agents of a Galaxy's Evolution (SAGE) Multiband Imaging Photometer (MIPS) 24um & 70um data, Chandra Supernova Remnants Catalog, and ATCA 4.8GHz continuum images of Dickel et al. (2005). Among 47 SNRs in the LMC, 7 SNRs show associated FIR emission. We present multi-wavelength view of 5 SNRs; DEM L249, N49, N63A, N132D, and the SNR in N4. N49 and N132D show morphological correlation in FIR and X-ray, suggesting that the FIR emission is from dust grains collisionally heated by X-ray emitting plasma. The FIR emission of N63A resembles H-alpha emission, which implies that the FIR line radiation could be dominant. The FIR images of the rest two objects, DEM L249 and SNR in N4, show no correlation to the other-waveband images.

• Journal of The Korean Astronomical Society
• /
• v.41 no.1
• /
• pp.7-15
• /
• 2008

We carried out a set of simulations to reproduce the performance of wide-field NEO surveys based on the revised population model of Near Earth Objects (NEOs) constructed by Morbidelli (2006). This is the first time where the new model is carefully compared with discovery statistics, and with the exception of population model, the simulation is identical to the procedure described in Moon et al. (2008). Our simulations show rather large discrepancy between the number of NEO discoveries made by the actual and the simulated surveys. First of all, unlike Bottke et al. (2002)'s, Morbidelli (2006)'s population model overestimates the number of NEOs. However, the latter reproduces orbit distributions of the actual population better. Our analysis suggests that both models significantly underestimate Amors, while overestimating the number of Apollos. Our simulation result implies that substantial modifications of both models are needed for more accurate reproduction of survey observations. We also identify Hungaria region (HU) to be one of the most convincing candidates that supply a large fraction of asteroids to the inner Solar System.