• Journal of KIISE
• /
• v.43 no.9
• /
• pp.1034-1042
• /
• 2016

Together with machine-learning frameworks, area-based imagery registration techniques can be easily applied to diverse types of image pairs without predefined features and feature descriptors. However, feature detectors are often used to quickly identify candidate image patch pairs, limiting the applicability of these registration techniques. In this paper, we propose a ConvNet (Convolutional Network) "Dart" that provides not only the matching metric between patches, but also information about their distance, which are helpful in reducing the search space of the corresponding patch pairs. In addition, we propose a ConvNet "Fad" to identify the patches that are difficult for Dart to improve the accuracy of registration. These two networks were successfully implemented using Deep Learning with the help of a number of training instances generated from a few registered image pairs, and were successfully applied to solve a simple image registration problem, suggesting that this line of research is promising.

• The Journal of The Korea Institute of Intelligent Transport Systems
• /
• v.9 no.1
• /
• pp.121-130
• /
• 2010

As the development of wireless environment and portable devices, navigation companies need more spatial information for Location Based Service and so they are investing heavily in this part. But my country have constructed lots of spatial dataset during the NGIS project, and then we can reduce costs by using constructed spatial dataset. The integration method of each building information between e-AIS and KLIS-rn is linked to designed mapping table by main building numbers and auxiliary cadastral number. e-AIS and KLIS-rn are integrated by the feature matching between CAD data and KLIS-rn. Due to the limitation of test data, more accuracy test will be required with various data sets.

• Journal of the Institute of Electronics Engineers of Korea SP
• /
• v.41 no.2
• /
• pp.11-22
• /
• 2004

Conventional algorithms of separating overlapped objects are mostly based on template matching methods and thus their application domain is restricted to 2D objects and the processing time increases when the number of templates (object models) does. To solve these problems, this paper proposes a new approach of separating the occluding object from the stack of 3D objects using the relationship between surfaces without any information on the objects. The proposed algorithm considers an object as a combination of surfaces which are consisted with a set of boundary edges. Overlap of 3D objects appears as overlap of surfaces and thus as crossings of edges in 2D image. Based on this observation, the types of edge crossings are classified from which the types of overlap of 3D objects can be identified. The relationships between surfaces are represented by an attributed graph where the types of overlaps are represented by relation values. Using the relation values, the surfaces pertained to the same object are discerned and the overlapping object on the top of the stack can be separated. The performance of the proposed algorithm has been proved by the experiments using the overlapped images of 3D objects selected among the standard industrial parts.

• Korean Journal of Social Welfare
• /
• v.65 no.1
• /
• pp.201-226
• /
• 2013

This study aims to identify changes in early work career of youth cohort entering the labor market pre and post-the economic crisis and compare career pathway types of different cohorts. Labor market experiences of youth cohort were constructed by sequencing the number of organizations, kinds of jobs, the scale of the business, and type of employment. In addition, a holistic sequence was created by including complementary factors. In this sense, the labor market experience in this study was conceptualized as a process involving continuous sequences and hierarchical and orderly changes which differs from a simple job mobility. Sequence analysis involving Optimal Matching method was conducted to examine whether such cohort-differences in labor market experiences were related to differences in distribution of career pathway types. The result showed that the post-economic crisis cohort had a relatively higher likelihood of falling into the non-employment type, unemployment type, non-corporate employment type, irregular employment type, and mobile employment type. These findings provide empirical evidence supporting the hypothesis that the employment precariousness of cohort has exacerbated after the economic crisis.

• Proceedings of the SCSK Conference
• /
• 2003.09b
• /
• pp.370-378
• /
• 2003

MFD (Make up Film Deterioration) is a gradual deterioration of applied make up and is a common problem experienced by most foundation users. Our investigation revealed that for 64% of all make up users MFD is their greatest consern is using foundations. Known that the primary cause of MFD is sebum secretion. We observed that the length of time prior to onset of MFD in people who produce high level of sebum varies significantly from person to person. This suggests that other factors besides quantity of sebum production can affect MFD. Control over this factor would, we believe, be key to developing longer-lasting makeups. We studied the relationship between MFD and skin surface conditions. Our study revealed that furrows on the skin surface affect MFD significantly. Sebum reaches the skin surface from sebaceous glands and flows along furrow on the skin. If there are many deep furrows, it takes longer for sebum to overflow. But if the furrows are few or shallow, sebum quickly overflows and spreads over the skin surface where it can degrade the make up film. Therefore even when the volume of sebum produced is the same, the rate of MFD will be different depending on the number and shape of the furrows. A longer-lasting foundation could be produced by matching personal skin condition, but this would be very difficult because individual variations in texture are very large. Therefore we approached the problem by attempting to impose sebum resistance in under make up and foundation. We have developed two new materials and make up products based on our theory. A new fluoroalkyl acrylate-methacrylates copolymer designed for incorporation in under make up is extremely sebum resistant and sweat proof. Another new acrylate polymer designed for inclusion in foundation absorbs sebum and changes to a solid. Usage tests confirm it is possible to reduce MFD by using under make up and foundation which incorporate our new materials to cover where skin furrows are few or shallow

• Proceedings of the IEEK Conference
• /
• 2000.07b
• /
• pp.885-888
• /
• 2000

Among the previous searching methods, there are the typical methods such as full search and three-step search, etc. Block motion estimation using exhaustive search is too computationally intensive. To apply in practice, recently proposed fast algorithms have been focused on reducing the computational complexity by limiting the number of searching points. According to the reduction of searching points, the quality performance is aggravated in those algorithms. In this paper, We present a fast and efficient search algorithm for block motion estimation that produces better quality performance and less computational time compared with a three-step search (TSS). Previously the proposed Two Step Search Algorithm (TWSS) by Fang-Hsuan Cheng and San-Nan sun is based on the ideas of dithering pattern for pixel decimation using a part of a block pixels for BMA (Block Matching Algorithm) and multi-candidate to compensate quality performance with several locations. This method has good quality performance at slow moving images, but has bad quality performance at fast moving images. To resolve this problem, the proposed algorithm in this paper considers spatial and temporal correlation using neighbor and previous blocks to improve quality performance. This performance uses neighbor motion vectors and previous motion vectors in addition, thus it needs more searching points. To compensate this weakness, the proposed algorithm uses statistical character of dithering matrix. The proposed algorithm is superior to TWSS in quality performance and has similar computational complexity

• Journal of the Institute of Electronics Engineers of Korea SP
• /
• v.45 no.6
• /
• pp.1-8
• /
• 2008

This paper introduces three efficient frame selection schemes to reduce the computation complexity for the multi-reference and variable block size Motion Estimation (ME). The proposed RSP (Reference Selection Pass) scheme can minimize the overhead of frame selection. The MFS (Modified Frame Selection) scheme can reduce the number of search points about 18% compared with existing schemes considering the motion of image during the reference frame selection process. In addition, the TPRFS (Two Pass Reference frame Selection) scheme can minimize the frame selection operation for the variable block size ME in H.264/AVC using the character of selected reference frame according to the block size. The simulation results show the proposed schemes can save up to 50% of the ME computation without degradation of image Qualify. Because the proposed schemes can be separated from the block matching process, they can be used with any existing single reference fast search algorithms.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.31 no.2B
• /
• pp.71-79
• /
• 2006

Address lookup is an essential function in the Internet routers, and it determines overall router performance. In this paper, we have thoroughly investigated the binary-search-based address lookup algorithms and proposed a new algorithm based on binary search on prefix lengths. Most of the existing binary search schemes perform binary search on prefix values, and hence the lookup speed is proportional to the length of prefixes or the log function of the number of prefixes. The previous algorithm based on binary search on prefix lengths has superior lookup performance than others. However, the algorithm requires very complicated pre-computation of markers and best matching prefixes in internal nodes since naive binary search is not possible in their scheme. This complicated pre-computation makes the composition of the routing table and incremental update very difficult. By using leaf-pushing, the proposed algorithm in this paper removes the complicated pre-computation of the Previous work in performing the binary search on prefix lengths. The performance evaluation results show that the proposed scheme has very good performance in lookup speed compared with previous works.

• Journal of Technologic Dentistry
• /
• v.34 no.3
• /
• pp.213-220
• /
• 2012

Purpose: The aim of this study was to determine the repeatability and reproducibility of two dental scanners. Methods: The master die and the stone replicas(Kavo, Germany) were digitized in touch-probe scanner(Incise, Renishaw, UK), white light scanner(Identica, Medit, Korea) to create 3-dimensional surface-models. The number of points in the point clouds from each reading were calculated and used as the CAD reference model(CRM). Discrepancies between the points in the 3-dimensional surface models and the corresponding CRM were measured by a matching-software(Power-Inspect R2, Delcam Plc, UK). The t-student test for one samples were used for statistical analysis. Results: The reproducibility of both scanner was within $3{\mu}m$, based on mean value. The mean value between measurements made directly on the touch probe scanner digital models and those made on the white light scanner digital models was $2.20-2.90{\mu}m$, and was statistically significant(P<0.05). Conclusion: With respect to adequate data acquisition, the reproducibility of dental scanner differs. Three-dimensional analysis can be applied to differential quality analysis of the manufacturing process as well as to evaluation of different analysis methods.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.16 no.5
• /
• pp.892-898
• /
• 2012

Rough classification methods are used to improving the recognition speed in many character recognition problems. In this case, some irreversible result can occur by an error in rough classification. Methods for duplicating each model in several classes are used in order to reduce this risk. But the errors by rough classfication can not be completely ruled out by these methods. In this paper, an recognition method is proposed to increase speed that matches models selectively without any increase in error. This method constructs a model graph using similarity between models. Then a search process begins from a particular point in the model graph. In this process, matching of unnecessary models are reduced that are not similar to the input pattern. In this paper, the proposed method is applied to the recognition problem of handwriting numbers and upper/lower cases of English alphabets. In the experiments, the proposed method was compared with the basic method that matches all models with input pattern. As a result, the same recognition rate, which has shown as the basic method, was obtained by controlling the out-degree of the model graph and the number of maintaining candidates during the search process thereby being increased the recognition speed to 2.45 times.