• Journal of Information Technology Applications and Management
• /
• v.23 no.2
• /
• pp.11-28
• /
• 2016

In this paper, we proposed a dictionary extension and a ternary decomposition technique to improve the effectiveness of Khmer word segmentation. Most word segmentation approaches depend on a dictionary. However, the dictionary being used is not fully reliable and cannot cover all the words of the Khmer language. This causes an issue of unknown words or out-of-vocabulary words. Our approach is to extend the original dictionary to be more reliable with new words. In addition, we use ternary decomposition for the segmentation process. In this research, we also introduced the invisible space of the Khmer Unicode (char\u200B) in order to segment our training corpus. With our segmentation algorithm, based on ternary decomposition and invisible space, we can extract new words from our training text and then input the new words into the dictionary. We used an extended wordlist and a segmentation algorithm regardless of the invisible space to test an unannotated text. Our results remarkably outperformed other approaches. We have achieved 88.8%, 91.8% and 90.6% rates of precision, recall and F-measurement.

• Journal of the Korean Institute of Telematics and Electronics
• /
• v.26 no.1
• /
• pp.10-19
• /
• 1989

This paper proposes the object modeling algorithm using depth data. In order to extract the shape of the object, surface is classified by the magnitude and the direction of Z gradient within the 2x2 mask from input depth data. For the object combined with several primitives is separated by kernel points. Further, spatial relationship between surface regions is extracted for the recognition. Finally, the effectiveness of this algorithm is demonstrated by several experiments.

• Proceedings of the IEEK Conference
• /
• 1999.06a
• /
• pp.857-860
• /
• 1999

It is made possible a removal of the preamble for carrier recovery and symbol-timing recovery by storing a burst in memory with low overhead QPSK demodulation and this demodulation method also effects frame efficiency improved by processed synchronization performance. In this paper, we have proposed that new algorithm for arc-tangent look-up table which transform the input I, Q data by phase. This I, Q data plays an important role in demodulation and makes demodulator with low-overhead by storing a burst in memory. To evaluate proposed new algorithm and symbol-timing recovery method, function simulation and timing verification have been done by using synopsys VHDL tool.

• Journal of Korea Society of Digital Industry and Information Management
• /
• v.9 no.2
• /
• pp.17-22
• /
• 2013

In this paper, we propose a concrete crack detection method using shape properties. It is based on morphology algorithm and crack features. We assume that an input image is contaminated by various noises. Thus, we use a morphology operator and extract patterns of crack. It segments cracks and background using opening and closing operations. Morphology based segmentation is better than existing integration methods using subtraction in detecting a crack it has small width. Also, it is robust to noisy environment. The proposed algorithm classifies the segmented image into crack and background using shape properties of crack. This method calculates values of properties such as the number of pixels and the maximum length of the segmented region. Also, pixel counts of clusters are considered. We decide whether the segmented region belongs to cracks according to those data. Experimental results show that our proposed crack detection method has better results than those by existing detection methods.

• Journal of The Institute of Information and Telecommunication Facilities Engineering
• /
• v.9 no.3
• /
• pp.80-85
• /
• 2010

In this paper, we present an algorithm for efficient control of backlight LED in order to reduce the power consumption of LCD TV. In addition, a pragmatic implementation method is presented. In conventional local dimming control, backlight LEDs of LCD TV are divided into $M{\times}N$ blocks. And, the proper luminance for each block is computed according to input image and used for dimming control. But, LED light of each block is diffused and affects the neighboring blocks, so that luminance of each block becomes larger than that of intent. In the proposed algorithm, dimming control values are reduced by the amount of quantity affected by the neighboring blocks using light spread function, resulting in additional reduction of power consumption.

• Proceedings of the Korean Institute of Intelligent Systems Conference
• /
• 2008.04a
• /
• pp.400-402
• /
• 2008

The human-computer interaction technology (HCI) that has played an important role in the exchange of information between human being and computer belongs to a key field for information technology. Recently, control studies through which robots and control devices are controlled by using the movements of a person's body or hands without using conventional input devices such as keyboard and mouse, have been going only in diverse aspects, and their importance has been steadily increasing. This study is proposing a recognition method of user's gestures by applying measurements from an acceleration sensor to the PCA algorithm.

• Proceedings of the Korean Society of Broadcast Engineers Conference
• /
• 1995.06a
• /
• pp.137-141
• /
• 1995

MASF is a kind of temporal filter proposed for noise reduction and temporal band limitation. MASF uses motion vectors to extract temporal information in spatial domain. Therefore, inaccurate motion information causes some distortions in MASF operation. Currently, bilinear interpolation after MBA(Block Matching Algorithm) is used for the motion estimation sheme of MASF. But, this method results in unreliable estimation when the object in image sequence has larger movement than the maximum displacement assumed in BMA or the input images are severely corrupted with noise. In order to solve this problem, we propose a hierarchical motion estimation algorithm for MASF. Experimental results show that the proposed method produces reliable output under large motion and noisy situations.

• Journal of Sensor Science and Technology
• /
• v.21 no.1
• /
• pp.21-27
• /
• 2012

In this study, we designed an embedded system for automatic condensation control(ESCC) of the car. This system heats the car glasses as and when it is needed that makes driving safer and convenient. The system was built on an ATmega128L central processing unit(CPU), using high-performance electrically erasable programmable read-only memory(EEPROM) complex programmable logic device(CPLD) ATF1504AS, using which an ESCC algorithm has been proposed. The source code was written in C language. The algorithm of work was written using the dew-point table. This system not only clears the condensation on the glass but also averts condensation. The designed ESCC system begins working once the input information comes close to the dew-point table information. This device enables a wider field of view, thereby increasing safety.

• Communications for Statistical Applications and Methods
• /
• v.9 no.1
• /
• pp.155-166
• /
• 2002

Neural networks have been studied as a popular tool for classification and they are very flexible. Also, they are used for many applications of pattern classification and pattern recognition. This paper focuses on Bayesian approach to feed-forward neural networks with single hidden layer of units with logistic activation. In this model, we are interested in deciding the number of nodes of neural network model with p input units, one hidden layer with m hidden nodes and one output unit in Bayesian setup for fixed m. Here, we use the latent variable into the prior of the coefficient regression, and we introduce the 'sequential step' which is based on the idea of the data augmentation by Tanner and Wong(1787). The MCMC method(Gibbs sampler and Metropolish algorithm) can be used to overcome the complicated Bayesian computation. Finally, a proposed method is applied to a simulated data.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.22 no.1
• /
• pp.162-172
• /
• 1997

In this paper, we investigated the performance of a binary feedback switch algorithm for the ABR(Available Bit Rate) service in ATM networks. A binary feedback switch is also called EFCI(Explicit Forward Congestion Indication) switch and can be classificed into input cell processing(IP) scheme according to processing methods for the EFCI bit in data-cell header. We proposed two implementation methods for the binary feedback switch according to EFCI-bit processing schemes, and analyzed the ACR(Allowed Cell Rate) of source and the queue length of switch for each scheme in steady state. In addition, we derived the upper and lower bounds for maximum and minimum queue lengths, respectively, and investigated the impact of ABR parameters on the queue length.