• Journal of Broadcast Engineering
• /
• v.12 no.4
• /
• pp.303-310
• /
• 2007

Motion Estimation (ME) is a core part of most Video compression systems since it affects directly the output video quality and the encoding time. The most basic method of ME, Full Search (FS) gives the highest visual quality but also has the problem of significant computational load. To solve this problem, many fast algorithm has been proposed. Among them, MVFAST and PMVFAST show impressive results in video quality and the computational load by using the correlation between motion vectors of adjacent blocks. In particular, PMVFAST reduces search points dramatically and also gives very high video quality by using the median predictor. In this paper, we propose a new algorithm that uses the redefined median predictor which reduces the number of search points and yields a high visual quality by reducing the number of thresholds and early termination conditions.

• Journal of the Korean Society of Surveying, Geodesy, Photogrammetry and Cartography
• /
• v.35 no.6
• /
• pp.441-452
• /
• 2017

Drones are increasingly being used to investigate disaster damage because they can quickly capture images in the air. It is necessary to extract the damaged area by registering the drones and the existing ortho-images in order to investigate the disaster damage. In this process, we might be faced the problem of registering two images with different time and spatial resolution. In order to solve this problem, we propose a new methodology that performs initial image transformation using line pairs extracted from images and association matrix, and final registration of images using linear features to refine the initial transformed result. The applicability of the newly proposed methodology in this study was evaluated through experiments using artifacts and the natural terrain areas. Experimental results showed that the root mean square error of artifacts and the natural terrain was 1.29 pixels and 4.12 pixels, respectively, and relatively high accuracy was obtained in the region with artifacts extracted a lot of linear information.

• The Journal of the Korea institute of electronic communication sciences
• /
• v.14 no.6
• /
• pp.1161-1170
• /
• 2019

In this study, we propose a model for predicting Cochlodinium polykrikoides red tide occurrence using deep neural networks. A deep neural network with eight hidden layers was constructed to predict red tide occurrence. The 59 marine and meteorological factors were extracted and used for neural network model training using satellite reanalysis data and meteorological model data. The red tide occurred in the entire dataset is very small compared to the case of no red tide, resulting in an unbalanced data problem. In this study, we applied over sampling with adding noise based data augmentation to solve this problem. As a result of evaluating the accuracy of the model using test data, the accuracy was about 97%.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.21 no.7
• /
• pp.1371-1378
• /
• 2017

In this paper, we present an upsampling technique for depth map image using selective bilateral weights and a color weight using laplacian function. These techniques prevent color texture copy problem, which problem appears in existing upsamplers uses bilateral weight. First, we construct a high-resolution image using the bicubic interpolation technique. Next, we detect a color texture region using pixel value differences of depth and color image. If an interpolated pixel belongs to the color texture edge region, we calculate weighting values of spatial and depth in $3{\times}3$ neighboring pixels and compute the cost value to determine the boundary pixel value. Otherwise we use color weight instead of depth weight. Finally, the pixel value having minimum cost is determined as the pixel value of the high-resolution depth image. Simulation results show that the proposed algorithm achieves good performance in terns of PSNR comparison and subjective visual quality.

• Transactions of the Korean Society of Mechanical Engineers B
• /
• v.41 no.7
• /
• pp.455-462
• /
• 2017

The magnetic interaction between elliptic Janus magnetic particles are investigated using a direct simulation method. Each particle is a one-to-one mixture of paramagnetic and nonmagnetic materials. The fluid is assumed to be incompressible Newtonian and nonmagnetic. A uniform magnetic field is applied externally in a horizontal direction. A finite-element-based fictitious domain method is employed to solve the magnetic particulate flow in the creeping flow regime. In the magnetic problem, the magnetic field in the entire domain, including the particles and the fluid, is obtained by solving the governing equation for the magnetic potential. Then, the magnetic forces acting on the particles are calculated via a Maxwell stress tensor formulation. In a single particle problem, it is found that the orientation angle at equilibrium is affected by the aspect ratio of the particle. As for the two-particle interaction, the dynamics and the final conformation of the particles are significantly influenced by the aspect ratio, the orientation, and the spatial positions of the particles. For the given positions of the particles, the fluid flow is also influenced by the orientation of each particle. The self-assembly structure of the particles is not a fixed one, but it varies with the above-mentioned factors.

• Journal of the Korean Society of Environmental Restoration Technology
• /
• v.4 no.2
• /
• pp.52-61
• /
• 2001

At least, there are intense pressures on the natural habitats from various disturbance, including urbanization, extension of industrial area, and road construction. These human land use result in fragmentation of landscape and natural habitat. The ecological consequences of habitat fragmentation include the direct effects of habitat loss and the indirect effects of reduced inter-patch dispersal. The decline of biological diversity has been rapidly declined by the habitat loss and fragmentation. Conservation strategists should consider not only the habitat amount of that must be preserved, but also the spatial configuration of habitat across the landscape. But, the paucity of available data for most species forces landscape ecologists to develop the concept, model, and theory. The developed theories are often misused in academic papers and practical applications. The development history, presumption of concept, model, theory is ignored. This tendency have leaded to failure of landscape restoration and the use of theory in conservation practice have come under increasing attacks. This paper will highlight the ecological theory that have proven the most influential in landscape ecology, restoration and conservation : the theory of island biogeography, the theory of nested subset analysis, minimum viable population(MVP), the theory of metapopulation dynamics. And, it find the problem and usefulness of four theory in application to real world. Consequently, the understanding of theoretical implication about landscape ecological theory is required. We must carefully apply the theory after examining the problem and availability of various theory because of no existence of only one general theory.

• Journal of the Korean association of regional geographers
• /
• v.12 no.6
• /
• pp.671-692
• /
• 2006

In recent years, it seems that a considerable stratum of working poor as a new urban poverty among labor population within large cities has been generated, deepening socio-spatial conflicts. This paper is to bring focus on the working poor which can be seen as a newly emerged special problem intermingling labor and poverty. It first begins with a consideration of definitions on the working poor, then discusses on the internal relationship between labor and poverty problems by analyzing causes of the generation of the working poor and looking into actual conditions of working poor from both static and dynamic perspectives, and finally considers existing policies for supporting the working poor, suggesting some alternative policy tasks for resolving the problem of the working poor.

• Journal of Korea Multimedia Society
• /
• v.7 no.4
• /
• pp.550-558
• /
• 2004

Previous researches on scalability problem of distributed virtual environment (DVE) have been mainly focused on spatial partitioning of area of interest (AOI). Congestion phenomena by avatar groups in AOI have been neglected relatively However, AOI congestion is highly related to scalability of DVE because it exhausts system resources such as network bandwidth and rendering time, and could be a bar to perform collaboration among participants. In this paper, this will be defined as the problem that must be solved for the realization of the scalable DVE, and a model will be proposed to measure and control congestion situation in AOI. The purposes of the proposed model are to prevent high density of participants in AOI, and to protect stable collaboration in DVE. For evaluation of the performance it is compared with a previous method by defining the resource cost model which is dynamically activated to AOI congestion.

• Journal of the Korean Society of Hazard Mitigation
• /
• v.2 no.1 s.4
• /
• pp.153-159
• /
• 2002

Disasters are the phenomena that we have to prevent the occurrence in order to keep the safety to human lives and properties, and if occurred, we have to minimize the economic, social, and mental costs of the occurred disasters or incidents. This research analyzes the optimal location of fire stations in terms of served population maximization in Seoul. This research introduces "the maximal covering location problem(MCLP)," one of the optimization techniques, as the primary research method. This research also applies a geographic information system into spatial distribution analyses of existing fire stations and observed fire incidents. Results from the analyses show that the existing location of fire stations and branches need to be improved. The dispatch location of fire engines should be reconsidered for rapid services of fire fighting.

• Journal of the Institute of Electronics Engineers of Korea SP
• /
• v.47 no.5
• /
• pp.89-99
• /
• 2010

A unified method for separating musical sources, singing voice for example, from stereophonic mixtures is provided. We usually have two observed signals in stereophonic music contents, where more than two instruments are played together. If we regard each instrument as source, this problem becomes an underdetermined source separation problem and cannot be solved by conventional methods, which infers the spatial environment of the downmixing process happens. Instead, source-specific information has been exploited to recover a particular instrumental source. This paper provides a unifying structure consists of heterogenious ad-hoc separate algorithms, which are designed for separating vocal sources using stereophonic channel information and dominant pitch information of the sources, respectively. Experiments on real world music contents show that the proposed unification can neutralize the drawbacks of the two ad-hoc separation algorithms and finally enhance the separation results.