• Journal of Intelligence and Information Systems
• /
• v.15 no.2
• /
• pp.49-59
• /
• 2009

Contents-based image retrieval methods are in general more objective and effective than text-based image retrieval algorithms since they use color and texture in search and avoid annotating all images for search. SIM(Self-organizing Image browsing Map) is one of contents-based image retrieval algorithms that uses only browsable mapping results obtained by SOM(Self Organizing Map). However, SOM may have an error in selecting the right BMU in learning phase if there are similar nodes with distorted color information due to the intensity of light or objects' movements in the image. Such images may be mapped into other grouping nodes thus the search rate could be decreased by this effect. In this paper, we propose an improved SIM that uses HSV color model in extracting image features with color quantization. In order to avoid unexpected learning error mentioned above, our SOM consists of two layers. In learning phase, SOM layer 1 has the color feature vectors as input. After learning SOM Layer 1, the connection weights of this layer become the input of SOM Layer 2 and re-learning occurs. With this multi-layered SOM learning, we can avoid mapping errors among similar nodes of different color information. In search, we put the query image vector into SOM layer 2 and select nodes of SOM layer 1 that connects with chosen BMU of SOM layer 2. In experiment, we verified that the proposed SIM was better than the original SIM and avoid mapping error effectively.

• Journal of Navigation and Port Research
• /
• v.29 no.5 s.101
• /
• pp.403-408
• /
• 2005

The techniques of estimating angle of arrival(AOA) have played a key role for enhancement of wireless communications using array antennas. Among those techniques, the superresolution algorithms, such as MUSIC and ESPRIT, calculate the covariance matrix of the array output vectors which are observed at the array antennas, and then by using eigen-decomposition of the covariance matrix, they estimate AOAs of the received signals with high accuracy. However, superresolution algorithms based eigenvalue decomposition fails to estimate AOAs under multipath environments. Under multipath environments, it is difficult to estimate AOAs of the received signals due to coherency and high-correlation. To resolve coherent signals, the covariance matrix is calculated by using the conventional spatial smoothing technique, and then the techniques based on eigen-descomposition is applied. The result of the conventional spatial smoothing technique, however, is obtained at the cost of losing effective spatial aperture. Moreover, the conventional technique ignores any information in the cross-correlations of the array outputs the subarrays. As the result, the performance for AOA estimation is degraded. In this paper, we propose a new spatial smoothing technique, which consider the cross-correlation for subarrays. By computer simulation, the AOA estimation performance of the proposed method is compared with the conventional method and evaluated.

• Journal of KIISE:Computer Systems and Theory
• /
• v.32 no.7
• /
• pp.349-358
• /
• 2005

False sharing is a result of co-location of unrelated data in the same unit of memory coherency, and is one source of unnecessary overhead being of no help to keep the memory coherency in multiprocessor systems. Moreover. the damage caused by false sharing becomes large in proportion to the granularity of memory coherency. To reduce false sharing in a page-based DSM system, it is necessary to allocate unrelated data objects that have different access patterns into the separate shared pages. In this paper we propose call-site tracing-based shared memory allocator. shortly CSTallocator. CSTallocator expects that the data objects requested from the different call-sites may have different access patterns in the future. So CSTailocator places each data object requested from the different call-sites into the separate shared pages, and consequently data objects that have the same call-site are likely to get together into the same shared pages. We use execution-driven simulation of real parallel applications to evaluate the effectiveness of our CSTallocator. Our observations show that by using CSTallocator a considerable amount of false sharing misses can be additionally reduced in comparison with the existing techniques.

• KIPS Transactions on Computer and Communication Systems
• /
• v.5 no.6
• /
• pp.135-142
• /
• 2016

We proposed the new method to estimate the blood pressure with the differential value of the digital arterial pulse waveform and BP relation equation. To get the digital arterial pulse waveform, we use the arterial pulse waveform measurement system that has digital air-pressure sensor device and smart phone. The acquired digital arterial pulse waveforms are classified as hypertension group, normal group, and hypotension group, and we can derive the average differential value between the highest point and lowest point of a single waveform of individuals along with the group. In this study, we found the functional correlation between the blood pressure and differential value as a form of BP relation equation through the regression process on the average of differential value and blood pressure value from a tonometer. The Experimental results show the BP relation equation can give easy blood pressure estimation method with a high accuracy. Although this estimation method has over 66 % error rate and does not give the high level of the accuracy for the diastolic compares to the commercial tonometer, the estimation results for the systolic show the high accuracy that has less than 10 % error rate.

• KIPS Transactions on Computer and Communication Systems
• /
• v.7 no.6
• /
• pp.145-154
• /
• 2018

A loosely coupled message broker system is a popular method for integrating distributed software components. Especially, a distributed broker structure with multiple brokers with active-standby or active-active message replicas are used to enhance availability as well as message processing performance. However, there are problems in both active-standby and active-active replica structure. The active-standby has relatively low processing performance and The active-active structure requires a high synchronization overhead. In this paper, we propose an active-active structure of replicas to increase the availability of the brokers without compromising its high fault-tolerancy. In the proposed structure, standby replicas process the requests of the active replicas so that load balancing is achieved without additional brokers, while the distributed coordinators are used for the synchronization process to decrease the overhead. We formulated the overhead incurred when synchronizing messages among replicas, and the formulation was used to support the experiment results. From the experiment, we observed that replicas of the active-active structure show better performance than the active stand-by structure with increasing number of users.

• Proceedings of the Korean Institute of Surface Engineering Conference
• /
• 2003.10a
• /
• pp.134-135
• /
• 2003

The dental implant materials required good mechanical properties, such as fatigue strength, combined with a high resistance to corrosion. For increasing fatigue resistance and delaying onset of stress corrosion cracking, shot peening has been used for > 50 years to extend service life of metal components. However, there is no information on the electrochemical behavior of shot peened and hydroxyapatite(HA) coated Ti-6Al-4V alloys. To increase fatigue strength, good corrosion resistance, and biocompatibility, the electrochemical characteristics of Ti/TiN/HA coated and shot peened Ti-6Al-4V alloys by electron beam physical vapor deposition(EB-PVD) have been researched by various electrochemical method in 0.9%NaCl. Ti-6Al-4V alloys were prepared under the condition of hydrogen and vacuum arc furnace. The produced materials were quenched at 1000$^{\circ}C$ under high purity dried Ar atmosphere and were hold at 500$^{\circ}C$ for 2 hrs to achieve the fatigue strength(1140㎫) of materials. Ti-6Al-4V alloys were prepared under the condition of hydrogen and vacuum arc furnace. Shot peening(SP) and sand blasting treatment was carried out for 1, 5, and 10min. On the surface of Ti-6Al-4V alloys using the steel balls of 0.5mm and alumina sand of 40$\mu\textrm{m}$ size. Ti/TiN/HA multilayer coatings were carried out by using electron-beam deposition method(EB-PVD) as shown Fig. 1. Bulk Ti, powder TiN and hydroxyapatite were used as the source of the deposition materials. Electrons were accelerated by high voltage of 4.2kV with 80 - 120mA on the deposition materials at 350$^{\circ}C$ in 2.0 X 10-6 torr vacuum. Ti/TiN/HA multilayer coated surfaces and layers were investigated by SEM and XRD. A saturated calomel electrode as a reference electrode, and high density carbon electrode as a counter electrode, were set according to ASTM GS-87. The potentials were controlled at a scan rate of 100 mV/min. by a potentiostat (EG&G Co.273A) connected to a computer system. Electrochemical tests were used to investigate the electrochemical characteristics of Ti/TiN/HA coated and shot peened materials in 0.9% NaCl solution at 36.5$^{\circ}C$. After each electrochemical measurement, the corrosion surface of each sample was investigated by SEM.

• The Transactions of the Korea Information Processing Society
• /
• v.6 no.10
• /
• pp.2587-2596
• /
• 1999

Recently, by the extension of skills like LAN, the wireless telecommunication network and the satelite services make it possible for the mobile computer users to access a database. A method to use the replicated database on a server to get new data without missing any information has been being studied. So far we have used the Static Replica Allocation(SRA) for the replication which is the method of the replication on the server. This method is to replicate the data on the replica server after a moving host is transferred to a cell. Since the network of the SRA is very good, and if there are few moving users, no trouble will happen. But if there is no moving users in a cell, the data will not be shared. Therefore, this paper is about the study of the method of relocation after replicating the data to the cells for the users(User Select Replica Allocation : USRA). We also analyze the access rate and the possibility which are closely related to the moving frequency of the mobile hosts and the numbers of the cells. As a result, We show that the 120% lower access cost and the 40%∼50% gains are achieved from the low mobility

• Journal of Intelligence and Information Systems
• /
• v.13 no.1
• /
• pp.15-27
• /
• 2007

In Korea, hepatoma is the thirdly frequent cause of death from cancer occupying 17.2% among the whole deaths from cancer and the rate of death from hepatoma comes to about 21's persons per one-hundred thousand ones. This paper proposes an automatic method for the extraction of areas being suspicious as hepatoma from a CT scan and evaluates the availability as an auxiliary tool for the diagnosis of hepatoma. For detecting tumors in the internal of the liver from CT scans, first, an area of the liver is extracted from about $45{\sim}50's$ CT scans obtained by scanning in 2.5-mm intervals starting from the lower part of the chest. In the extraction of an area of the liver, after unconcerned areas outside of the ribs being removed, areas of the internal organs are separated and enlarged by using intensity information of the CT scan. The area of the liver is extracted among separated areas by using information on position and morphology of the liver. Since hepatoma is a hypervascular turner, the area corresponding to hepatoma appears more brightly than the surroundings in contrast-enhancement CT scans, and when hepatoma shows expansile growth, the area has a spherical shape. So, for the extraction of areas of hepatoma, areas being brighter than the surroundings and globe-shaped are selected as candidate ones in an area of the liver, and then, areas appearing at the same position in successive CT scans among the candidates are discriminated as hepatoma. For the performance evaluation of the proposed method, experiment results obtained by applying the proposed method to CT scans were compared with the diagnoses by radiologists. The evaluation results showed that all areas of the liver and liver tumors were extracted exactly and the proposed method has a high availability as an auxiliary diagnosis tools for the discrimination of liver tumors.

• The Journal of Engineering Geology
• /
• v.18 no.4
• /
• pp.381-391
• /
• 2008

The eigenvalue technique is introduced to overcome the problem of truncation errors caused by temporal discretization of numerical analysis. The eigenvalue technique is different from simulation in that only the space is discretized. The spatially discretized equation is diagonized and the linear dynamic system is then decoupled. The time integration can be done independently and continuously for any nodal point at any time. The results of eigenvalue technique are compared with the exact solution and FEM numerical solution. The eigenvalue technique is more efficient than the FEM to the computation time and the computer storage in the same conditions. This technique is applied to the solute transport analysis in nonuniform flow fields around underground storage caverns. This method can be very useful for time consuming simulations. So, a sensitivity analysis is carried out by using this method to analyze the safety of caverns from nearly located contaminant sources. According to the simulations, the reaching time from source to the nearest cavern may takes 50 years with longitudinal dispersivity of 50 m and transversal dispersivity of 5 m, respectively.

• Journal of Sensor Science and Technology
• /
• v.6 no.5
• /
• pp.346-355
• /
• 1997

In this paper, the analyses of the stress disturibution and frequency characteristics of silicon microstructures for an accelerometer were performed using the general purpose finite element simulation program, ANSYS. From the analyses, we determined the parameter values of a new 6-beam piezoresistive accelerometer applicable to the accelerometer's specification in airbag system of automobile. Then, the mass paddle radius, beam length, beam width, and beam thickness of the designed accelerometer were$500{\mu}m$, $350{\mu}m$, $100{\mu}m$, and $5{\mu}m$, respectively and two different seismic masses with 0.4 mg and 0.8 mg were defined on the same sensor structure. The designed 6- beam accelerometers were fabricated on the selectively diffused (111)-oriented $n/n^{+}/n$ silicon substrates and the characteristics of the fabricated accelerometers were investigated. Then, we used a micromachining technique using porous silicon etching method for the formation of the micromechanical structure of the accelerometer.