• 한국신재생에너지학회:학술대회논문집
• /
• 2009.11a
• /
• pp.349-349
• /
• 2009

High quality ZnO:Al films were prepared on glass substrates by in-line RF magnetron sputtering and their surface morphologies were modified by wet-etching process in dilute acid solution to improve optical properties for application to silicon thin film solar cells as front electrode. The as-deposited films show a strong preferred orientation in [001] direction under our experimental conditions. A low resistivity below $5{\times}10^{-4}{\Omega}{\cdot}cm$ and high optical transmittance above 80% in a visible range are achieved in the films deposited at optimized conditions. After wet-etching, the surface morphologies of the films are changed dramatically depending on the deposition conditions, especially working pressure. The optical properties such as total/diffuse transmittance, haze and angular resolved distribution of light are varied significantly with the surface morphology feature, whereas the electrical properties are seldom changed. The cell performances of silicon thin film solar cells fabricated on the textured films are also evaluated in detail with comparison of commercial $SnO_2$:F films.

• Journal of the Korean Society for Nondestructive Testing
• /
• v.29 no.3
• /
• pp.214-218
• /
• 2009

The occlusal force of the tooth leads to loss of tooth tissue owing to attrition and abrasion, and may cause abfraction and pathological change of the dentin. Thus, we developed finite element models, examined them by applying ordinary occlusal force, and analyzed the stress distribution. Specimens used were mandibular first premolars from 15 Korean males and 13 females and were made into finite element models from medical images that were obtained using a Micro-CT. We have found that the irregular feature of the tooth is not only useful to masticating and pronouncing as well known, but it is also suitable for protecting inner tissue by dispersing stress and delivering proper pressure to periodontal tissue to continue a physiological action. Also, image analysis could let us know the factor that is the cause of a disorder due to stress concentration in the cervical line. These results are expected to support the field of dental treatment planning, operating procedure and clinical trial, and the advance of technical expertise to develop implants and dentures.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.9 no.9
• /
• pp.3751-3770
• /
• 2015

Many sorts of image processing software facilitate image editing and also generate a great number of doctored images. Forensic technology emerges to detect the unintentional or malicious image operations. Most of forensic methods focus on the detection of single operations. However, a series of operations may be used to sequentially manipulate an image, which makes the operation detection problem complex. Forensic investigators always want to know as much exhaustive information about a suspicious image's entire processing history as possible. The detection of the operation chain, consisting of a series of operations, is a significant and challenging problem in the research field of forensics. In this paper, based on the histogram distribution uniformity of a manipulated image, we propose an operation chain detection scheme to identify histogram equalization (HE) followed by the dither-like operation (DLO). Two histogram features and a local spatial feature are utilized to further determine which DLO may have been applied. Both theoretical analysis and experimental results verify the effectiveness of our proposed scheme for both global and local scenarios.

• Proceedings of the Korean Society for Language and Information Conference
• /
• 2007.11a
• /
• pp.303-310
• /
• 2007

Named Entity Recognition (NER) is always limited by its lower recall resulting from the asymmetric data distribution where the NONE class dominates the entity classes. This paper presents an approach that exploits non-local information to improve the NER recall. Several kinds of non-local features encoding entity token occurrence, entity boundary and entity class are explored under Conditional Random Fields (CRFs) framework. Experiments on SIGHAN 2006 MSRA (CityU) corpus indicate that non-local features can effectively enhance the recall of the state-of-the-art NER systems. Incorporating the non-local features into the NER systems using local features alone, our best system achieves a 23.56% (25.26%) relative error reduction on the recall and 17.10% (11.36%) relative error reduction on the F1 score; the improved F1 score 89.38% (90.09%) is significantly superior to the best NER system with F1 of 86.51% (89.03%) participated in the closed track.

• Korea-Australia Rheology Journal
• /
• v.19 no.2
• /
• pp.89-95
• /
• 2007

The flow characteristics of chicken blood in a micro-tube with a $100{\mu}m$ diameter are investigated using a micro-Particle Image Velocimetry (PIV) technique. Chicken blood with 40% hematocrit is supplied into the micro-tube using a syringe pump. For comparison, the same experiments are repeated for human blood with 40% hematocrit. Chicken blood flow has a cell-free layer near the tube wall, and this layer's thickness increases with the increased flow speed due to radial migration. As a hemorheological feature, the aggregation index of chicken blood is about 50% less than that of human blood. Therefore, the non-Newtonian fluid features of chicken blood are not very remarkable compared with those of human blood. As the flow rate increases, the blunt velocity profile in the central region of the micro-tube sharpens, and the parabolicshaped shear stress distribution becomes to have a linear profile. The viscosity of both blood samples in a low shear rate condition is overestimated, while the viscosity in a high shear rate range is underestimated due to radial migration and the presence of a cell-depleted layer.

• Applied Microscopy
• /
• v.25 no.1
• /
• pp.111-121
• /
• 1995

The ultrastructure of spermatozoa in Hynobius leechii was examined with transmission electron microscope and compared with those of other urodeles and anurans. The perforatorium and the tail show most of the common feature of urodeles. However, there were several ultrastructural characteristics in an acrosome, nucleus, neck, axial rod and mitochondria related to the nucleus. The acrosome was trifoliate in transverse sections and the perforatorium consisted of two different concentric parts with a fine sharp point in a subacrosomal lumen. The nucleus consisted of two different regions of chromatin area and nuclear ridge. The nuclear ridge was composed of several bundles of five to ten minute tubular subunits of 19 nm diameter in this species, while in higher urodeles it was well developed in multi layers. The protoplasmic bead was separated by a cytoplasmic canal except the connection with only the distal portion of the nucleus. The neck was a short cylinder and contained pericentriolar material with transverse striations. In Hynobius the ring was not elongated to the tail and the mitochondria were distributed only in the protoplasmic bead around the nucleus as in Cryptobranchus, while in higher groups of urodeles it elongates to the length of the middle piece and the mitochondria follow the ring. The ring elongation may be related to the distribution of mitochondria. Hynobiidae and Cryptobranchidae are closely related based on structure of neck, the axial rod and location of mitochondria, although they are different in the composition of the axial rod.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.14 no.7
• /
• pp.3076-3092
• /
• 2020

An IKPCA-ELM-based intrusion detection method is developed to address the problem of the low accuracy and slow speed of intrusion detection caused by redundancies and high dimensions of data in the network. First, in order to reduce the effects of uneven sample distribution and sample attribute differences on the extraction of KPCA features, the sample attribute mean and mean square error are introduced into the Gaussian radial basis function and polynomial kernel function respectively, and the two improved kernel functions are combined to construct a hybrid kernel function. Second, an improved particle swarm optimization (IPSO) algorithm is proposed to determine the optimal hybrid kernel function for improved kernel principal component analysis (IKPCA). Finally, IKPCA is conducted to complete feature extraction, and an extreme learning machine (ELM) is applied to classify common attack type detection. The experimental results demonstrate the effectiveness of the constructed hybrid kernel function. Compared with other intrusion detection methods, IKPCA-ELM not only ensures high accuracy rates, but also reduces the detection time and false alarm rate, especially reducing the false alarm rate of small sample attacks.

• Journal of the Institute of Convergence Signal Processing
• /
• v.1 no.1
• /
• pp.49-57
• /
• 2000

In this study, a new algorithm for canceling a MRI artifact due to the translational motion In the image plane is described. Unlike the conventional iterative phase retrieval algorithm, in which there is no guarantee for the convergence, a direct method for estimating the motion is presented. In previous approaches, the motions in the x(read out) direction and the y(phase encoding) direction were estimated simultaneously. However, the feature of x and y directional motions are different from each other. By analyzing their features, each x and y directional motion is canceled by the different algorithms in two steps. First, it is noticed that the x directional motion corresponds to a shift of the x directional spectrum of the MRI signal, and the non-zero area of the spectrum just corresponds to the projected area of the density function on the x axis. So the motion is estimated by tracing the edges between non-zero area and zero area of the spectrum, and the x directional motion is canceled by shifting the spectrum in an reverse direction. Next, the y directional motion is canceled by using a new constraint condition, with which the motion component and the true image component can be separated. This algorithm is shown to be effective by using a phantom image with simulated motion.

• Journal of Broadcast Engineering
• /
• v.5 no.1
• /
• pp.31-40
• /
• 2000

In this paper, we propose an edge histogram to efficiently represent the edge distribution in the image for MPEG-7. To this end, we adopt global, semi-global, and local edge histogram bins. Also, we extract the edge information from the image in terms of image blocks rather than pixels, which reduces the extraction complexity and is also applicable to the block-based compression standards such as MPEG-1, and 2. Experimental results show that the proposed method yields better retrieval accuracy and feature extraction speed comparing to other non-homogeneous texture descriptors of MPEG-7 including the wavelet-based descriptor and local edge-based descriptor.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.10 no.12
• /
• pp.5464-5484
• /
• 2016

Few samples face recognition has become a highly challenging task due to the limitation of available labeled samples. As two popular paradigms in face image representation, sparse component analysis is highly robust while parts-based paradigm is particularly flexible. In this paper, we propose a probabilistic generative model to incorporate the strengths of the two paradigms for face representation. This model finds a common spatial partition for given images and simultaneously learns a sparse component analysis model for each part of the partition. The two procedures are built into a probabilistic generative model. Then we derive the score function (i.e. feature mapping) from the generative score space. A similarity measure is defined over the derived score function for few samples face recognition. This model is driven by data and specifically good at representing face images. The derived generative score function and similarity measure encode information hidden in the data distribution. To validate the effectiveness of the proposed method, we perform few samples face recognition on two face datasets. The results show its advantages.