• 한국정보통신학회:학술대회논문집
• /
• 한국해양정보통신학회 2010년도 춘계학술대회
• /
• pp.612-613
• /
• 2010

The common image processing method treat the image as the array of pixels, and every pixel with its eight neighbors who directly around it form a square neighbor domain, sometimes may be more than that such as $5{\times}5$, or $7{\times}7$, and then adopt the convolution and template for every possible pixel value and then divide a attenuation factor, restrict the result in the area between 0 and 255, while the original and processed value record respectively. During the whole procedure, the result sole exist and represent the processing without changing the original pixel.

• 대한원격탐사학회:학술대회논문집
• /
• 대한원격탐사학회 2003년도 Proceedings of ACRS 2003 ISRS
• /
• pp.153-155
• /
• 2003

Laser scanning has become a viable technique for the collection of a large amount of accurate 3D point data densely distributed on the scanned object surface. The inherent 3D nature of the sub-randomly distributed point cloud provides abundant spatial information. To explore valuable spatial information from laser scanned data becomes an active research topic, for instance extracting digital elevation model, building models, and vegetation volumes. The sub-randomly distributed point cloud should be segmented and classified before the extraction of spatial information. This paper investigates some exist segmentation methods, and then proposes an octree-based split-and-merge segmentation method to divide lidar data into clusters belonging to 3D planes. Therefore, the classification of lidar data can be performed based on the derived attributes of extracted 3D planes. The test results of both ground and airborne lidar data show the potential of applying this method to extract spatial features from lidar data.

• 대한전기학회:학술대회논문집
• /
• 대한전기학회 2004년도 심포지엄 논문집 정보 및 제어부문
• /
• pp.102-104
• /
• 2004

Finger language is a typical tool for deaf persons. But learning the finger language for non-handicapped persons is very difficult. To overcome these difficulties, a new communication method using visual function is developed recently. Even though the developed system uses the visual function, it needs expensive equipments such as camera and computer. To be used in the real environments, the cost of equipments is a critical factor. If the recognition system for the finger language can be developed with low price equipments, the system can be used in the notebook or cellular phone. The image captured by PC camera was processed by preprocessing algorithm. To recognize the finger language, the resulting image was divide into $5{\times}5$ sections. The recognition system uses a similarity method and position information. The simulation results shows the effectiveness of the proposed algorithm.

• 한국품질경영학회:학술대회논문집
• /
• 한국품질경영학회 2006년도 춘계학술대회
• /
• pp.100-105
• /
• 2006

Many corporations will get much mote benefits if they investigate the model deciding to purchasing mobile phone in that cell phone of our country has about 20% of market share in the world market. Considering this situation to divide cell phone market between CDMA and GSM in world market, it is not easy to make cell phone to satisfy the suitable method which each country wants. Through this study, we anticipate that it will offer the measuring method about consumer purchasing decision of each country. The research is to study how it is Effected by any factor when comsumer selects product and make research model. Based on the research model, we is to make a question. The research Is to complete the theoretical system on the basis of records about the cell phone and to study the improvement strategic using AHP and Conjoint Analysis which analyzes the consuming trend to customer's cell phone.

• 동의생리병리학회지
• /
• 제20권1호
• /
• pp.245-256
• /
• 2006

This research sought to survey oriental medicine diagnosis methods currently practiced, analyze their advantages and disadvantages, and work out clinical establish direction for oriental medicine diagnosis methods. Oriental medicine diagnosis methods currently practiced in the related circles are categorized into traditional Korean diagnosis methods and holistic auxiliary diagnosis methods. The traditional Korean diagnosis method focuses on treating diseases of traditional Korean health management methods which are based on the bodily self-viability capabilities according to the Orient's viewpoint of health. Under the diagnosis method, based on the cognition of maximizing the state of the bodily self-viability capabilities together with the characteristics of diseases, symptoms, pulse, first face-to-face patient observation, physical constitution, and life principle are managed according to form, color, pulse and symptom which divide the bodily viability capacities into inherent and acquired elements amid both elements interacting.

• 한국지진공학회:학술대회논문집
• /
• 한국지진공학회 2003년도 추계 학술발표회논문집
• /
• pp.3-7
• /
• 2003

The velocity structure beneath the CHNB broadband station is determined by receiver function analysis using by from teleseismic P waveforms. The detailed broadband receiver functions are obtained by stacking method for source-equalized vertical, radial and tangential components of teleseismic P waveforms. A time domain inversion uses the stacked radial receiver function to determine vertical P wave velocity structure beneath the station. The crustal velocity structures beneath the stations are estimated using the receiver function inversion method in the case at the crustal model parameterized by many thin, flat-lying, homogeneous layers. Events divide into 4 groups. four azimuths corresponding to events in group a(southwest), b(south), c(southeast), d(northeast). The result of crust at model inversion shows the crustal velocity structure beneath the CHNB station varies smoothly with increasing depth. The conard discontinuity lies around 18 km and moho discontinuity lies range from 30 to 34 km.

• 한국정보통신학회:학술대회논문집
• /
• 한국해양정보통신학회 2009년도 춘계학술대회
• /
• pp.156-159
• /
• 2009

Speckle noise reduction for 3D power doppler ventricle coherent image for restoration and enhancement using wavelet packet transform with separated thresholding is presented. Wavelet Packet Transform divide into low frequency component image to high frequency component image to be multi-resolved. speckle noise is located on high frequency component in multiresolution image mainly. A ventricle image is transformed and inversed with separated threshold function from low to high resolved images for restoration to be utilize visualization for ventricle diagnosis. The experimental result shows that the proposed method has better performance in comparison with the conventional method.

• ETRI Journal
• /
• 제43권3호
• /
• pp.538-548
• /
• 2021

This paper proposes an approach to improve the performance of no-reference video quality assessment for sports videos with dynamic motion scenes using an efficient spatiotemporal model. In the proposed method, we divide the video sequences into video blocks and apply a 3D shearlet transform that can efficiently extract primary spatiotemporal features to capture dynamic natural motion scene statistics from the incoming video blocks. The concatenation of a deep residual bidirectional gated recurrent neural network and logistic regression is used to learn the spatiotemporal correlation more robustly and predict the perceptual quality score. In addition, conditional video block-wise constraints are incorporated into the objective function to improve quality estimation performance for the entire video. The experimental results show that the proposed method extracts spatiotemporal motion information more effectively and predicts the video quality with higher accuracy than the conventional no-reference video quality assessment methods.

• Communications for Statistical Applications and Methods
• /
• 제30권3호
• /
• pp.343-354
• /
• 2023

Recently, there have been many improvements in general language models using architectures such as GPT-3 proposed by Brown et al. (2020). Nevertheless, training complex models can hardly be done if the number of data is very small. Data augmentation that addressed this problem was more than normal success in image data. Image augmentation technology significantly improves model performance without any additional data or architectural changes (Perez and Wang, 2017). However, applying this technique to textual data has many challenges because the noise to be added is veiled. Thus, we have developed a novel method for performing data augmentation on text data. We divide the data into signals with positive or negative meaning and noise without them, and then perform data augmentation using k-doc augmentation to randomly combine signals and noises from all data to generate new data.

• Journal of Power Electronics
• /
• 제18권5호
• /
• pp.1489-1500
• /
• 2018

Partial shading analysis of large-scale photovoltaic (PV) arrays has recently become a theoretically and numerically challenging issue, and it is necessary for PV system designers. The main contributions of this study are the following: 1) A PSIM-based macro-model was employed because it is remarkably fast, has high precision, and has no convergence issues. 2) Three types of equivalent macro-models were developed for the transformation of a small PV sub-array with uniform irradiance to a new macro-model. 3) On the basis of the proposed new macro-model, a tearing method was established, which can divide a large-scale PV array into several small sub-arrays to significantly improve the efficiency improvement of a simulation. 4) Three platforms, namely, PSIM, PSpice, and MATLAB, were applied to evaluate the proposed tearing method. The proposed models and methods were validated, and the value of this research was highlighted using an actual large-scale PV array with 2420 PV modules. Numerical simulation demonstrated that the tearing method can remarkably improve the simulation efficiency by approximately thousands of times, and the method obtained a precision of nearly 6.5%. It can provide a useful tool to design the optimal configuration of a PV array with a given shading pattern as much as possible.