• Journal of Korea Society of Digital Industry and Information Management
• /
• v.19 no.4
• /
• pp.127-136
• /
• 2023

Many research results on the traffic enforcement of illegal driving of twowheeled motor vehicles using license plate recognition are introduced. Deep learning convolutional neural networks can be used for character and word recognition of license plates because of better generalization capability compared to traditional Backpropagation neural networks. In the plates of twowheeled motor vehicles, the interdependent government and city words are included. If we implement the mutually independent word recognizers using error correction rules for two word recognition results, efficient license plate recognition results can be derived. The CPU based convolutional neural network without library under real time processing has an advantage of low cost real application compared to GPU based convolutional neural network with library. In this paper twowheeled motor vehicle license plate recognition algorithm is introduced using CPU based deep-learning convolutional neural network. The experimental results show that the proposed plate recognizer has 96.2% success rate for outdoor twowheeled motor vehicle images in real time.

• Journal of information and communication convergence engineering
• /
• v.22 no.2
• /
• pp.121-126
• /
• 2024

In enterprise environments, file owners are often required to share critical files with other users, with encryption-based file delivery systems used to maintain confidentiality. However, important information might be leaked if the cryptokey used for encryption is exposed. To recover confidentiality, the file owner must then re-encrypt and redistribute the file along with its new encryption key, which requires considerable resources. To address this, we propose a key management server that minimizes the distribution of encryption keys when critical files are compromised, with unique encryption keys assigned for each registered user to access critical files. While providing the targeted functions, the server employs a level of system resources comparable to that of legacy digital rights management. Thus, when implemented in an enterprise environment, the proposed server minimizes cryptokey redistribution while maintaining accessibility to critical files in the event of an information breach.

• 한국디지털정책학회:학술대회논문집
• /
• 2004.11a
• /
• pp.25-37
• /
• 2004

This paper compares long-run buy-and-hold returns of privatization initial public offerings to those of domestic stock markets of respective countries using a sample of 29 privatized telecom initial public offerings from 27 countries. The evidence indicates that the privatization IPOs significantly outperform their domestic stock markets if the returns are equally-weighted while they do not outperform the markets if value-weighted. In addition, this paper analyzes the cross-sectional determinants of long-run buy-and- hold returns of privatized telecom shares. The results indicate that the long-run performance of privatized telecom IPOs is moderately related to the proxies of policy uncertainty or systematic risk while the size of the firm and some market wide variables such as the accounting standard, origin of commercial law, and the corporate governance scheme significantly affect the stock performance of privatized telecom shares.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.25 no.12B
• /
• pp.2052-2059
• /
• 2000

Digital IF(Intermediate Frequency) 처리단과 같은 고속과 저전력을 요구하는 필터에서 덧셈기만을 사용하여 CSD(Canonical Signed Digit)형의 필터계수들을 구현하는 구조가 널리 연구되고 있다. 본 논문에서는 선형위상 FIR(Finite Impulse Response) 필터의 CSD형 필터계수들을 최소의 덧셈으로 구현할 수 있는 아키텍처를 제안한다. 1과 -1로 이루어진 필터계수 표에서 공통패턴을 공유함으로서 덧셈의 수를 줄이는 방법이 이미 연구되었다. 본 논문은 비트 shift, 비트 add, 비트 반전을 통하여 인공의 공통패턴을 만들어서 이미 존재하는 공통패턴에 합류시킴으로서 덧셈의 수를 더욱 줄일 수 있는 방법을 제안한다. CDMA 이동통신 단말기의 IF단에 사용되는 사양의 디지털 필터를 73탭의 CSD형 계수로 구현하여 9.2%의 덧셈 감소의 효과가 있음을 보였다.

• 제어로봇시스템학회:학술대회논문집
• /
• 2000.10a
• /
• pp.97-97
• /
• 2000

In this paper, a signal processing algorithm for software GPS receiver is proposed. The signal processor takes snapshot of the sampled If signal from the RF section of the GPS receiver. All the processing for code and carrier tracking and correlation are implemented using the digital signal processing techniques. In order to achieve fast code acquisition, correlation of the incoming GPS signal is performed using the FFT method, After code acquisition, to reduce the Doppler shift effect and increase the accuracy, the interpolation or the tracking are performed. The performance of the proposed processing algorithm is first evaluated using matlab/simulink. A signal acquisition board for sampling and logging GPS IF signal form the Mitel GPS RF chip set is constructed. In order to analyze the performance of the designed algorithm the experiments are performed and the results are analyzed.

• Proceedings of the KSRS Conference
• /
• 2007.10a
• /
• pp.434-437
• /
• 2007

Maps contain and effectively visualize a number of spatial information. Advances in GIS enable researchers to analyze and represent spatial information through digital maps. Choropleth maps represent different quantities showing usually rates, percentages or densities. Generally, researchers make choropleth maps using raw rates. But, if the events are rare, raw rates cannot be sufficient in representing spatial phenomena. That is to say, if the population is large and events are rare, we cannot be sure that the raw rate is correct. The objective of this study is to make choropleth maps by several rate calculation methods and compare them. We use three methods in choropleth mapping; a raw rate, empirical Bayesian method, and spatial rate method which use prior probabilities. The experiments reveal that maps are somewhat different by used methods. We suggest that a raw rate method can not be an only way to make a rate map and researchers should choose an appropriate method for their objectives.

• Proceedings of the KIPE Conference
• /
• 2018.07a
• /
• pp.332-333
• /
• 2018

This paper presents an advanced abnormal over-current protection with $SuperFET^{(R)}$ 800V MOSFET in Flyback converter. In advanced abnormal over-current protection, digital pattern generator is proposed to detect a steep di/dt current condition when secondary rectifier diode or the transformer is shorted. If current sensing signal is larger than current limit during consecutive switching cycle, Gate signal will be stopped for 7 internal switching periods. If the abnormal over-current maintains pattern, the controller goes into protection mode. The Advanced over-current protection has been implemented in a 0.35um BCDMOS process (ON Semiconductor process).

• 한국디지털정책학회:학술대회논문집
• /
• 2005.11a
• /
• pp.465-474
• /
• 2005

Knowledge has been widely recognized as a critical resource for competitive advantage of the organization. However, little has been done in the literature regarding under what conditions that organizations should develop knowledge internally or acquire knowledge from outside. Therefore, the purpose of this study is to examine the effects of some key factors such as environment and organizational climate, on the organization's decision to knowledge sourcing. Our findings suggest that organizations are more likely to acquire knowledge from outside if the environment is complex, munificent and dynamic; if they possess higher levels of goal-oriented and autonomy.

• Proceedings of the KIEE Conference
• /
• 1998.11b
• /
• pp.435-437
• /
• 1998

In this productive system, it needs to control the each axis motion harmoniously to perform accurately for the manufacturing, transporting and printing. Independent Axis Control usually used for this objection. However, if Independent Axis Control mismatched the parameter of each axis system or in the case of free curve tracking or the case of high speed control, there would be big contour error so that cannot achieve control objection. As a result, there is Contour Control method suggested to supply for this defect. This paper carried modeling of biaxial system and implemented Independent Axis Control & Contouring Control on straight line, circular, and coner path by simulation and experiment. If feedrate increased, contour error growed. In consequence, according to this factor, we introduced contouring controller, so we could find the fact that contour error was reduced more than that of independent axis control about each path.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.9 no.1
• /
• pp.11-17
• /
• 1984

A study is presented showing the effect of carrier phase error on the error-rate of a (Class I, II, IV)PR shaped SQAM digital communication system. A simple upper bound on the probabilty of error as a function of phase error is derived and compared to one another. The results show that the three system can be used appropriately if carrier phase error is less than 3$^{\circ}$. However, if phase error is larger than 3, the PR class I is the best choice.