• 한국정보과학회논문지:컴퓨팅의 실제 및 레터
• /
• 제16권12호
• /
• pp.1234-1238
• /
• 2010

최근 센서 네트워크를 이용하여 이동 객체의 정보를 모니터링 하는 응용에 대한 연구가 활발하게 이루어지고 있다. 네트워크 전체 영역에서 대상 객체가 원하는 밀도로 분포하는 영역을 찾아내는 밀도 질의는 객체 모니터링 응용의 한 분야이다. 본 논문에서는 에너지 효율적인 질의 처리를 위한 동종 센서 기반의 인-네트워크 밀도 질의 처리 기법을 제안한다. 제안하는 기법은 밀도 질의 처리의 정확도를 높이고, 에너지 소비를 최소화하기 위한 가능성 기반 예상 지역 선정 기법과 센싱 영역 면적 기반 결과 보정 기법을 수행한다. 제안하는 기법의 우수성을 보이기 위해 시뮬레이션을 통해 기존에 제안된 밀도 질의 처리 기법과의 성능을 비교하였다. 그 결과 기존 기법에 비해 질의 처리를 위한 에너지 소모는 약 92% 감소하였고, 그에 따른 네트워크 생존 시간이 증가하였다. 덧붙여, 기존 기법보다 제안하는 기법의 질의 결과가 더 높은 정확도를 보장한다.

• 한국산업정보학회논문지
• /
• 제22권2호
• /
• pp.61-70
• /
• 2017

본 논문에서는 CPU기반 병렬처리인 OpenMP를 적용하여 헤드라이트의 차종을 판별하기 위한 처리속도를 개선하고, 생산부품의 얼라인먼트를 보정 후 특징을 추출하여 차종을 판단하는 방법을 제안하였다. 제안방법의 세부성능을 분석하기 위해 그레이영상을 이용한 방법 및 라인 검출을 이용한 방법과의 판별성능과 처리속도를 비교 분석하였다. 분석결과, 판별성능의 경우 제안방법과 라인 검출을 이용한 방법이 좋은 성능을 보였지만 처리속도 측면에서 라인검출 방법보다 제안 방법이 더 좋은 성능을 보였다. 처리속도의 경우 그레이 영상을 이용한 방법이 가장 좋은 성능을 보였으나 판별성능 측면에서 제안방법이 그레이 영상을 이용한 방법보다 좋다고 할 수 있다.

• Journal of Information Processing Systems
• /
• 제14권6호
• /
• pp.1464-1479
• /
• 2018

In face recognition, sometimes the number of available training samples for single category is insufficient. Therefore, the performances of models trained by convolutional neural network are not ideal. The small sample face recognition algorithm based on novel Siamese network is proposed in this paper, which doesn't need rich samples for training. The algorithm designs and realizes a new Siamese network model, SiameseFacel, which uses pairs of face images as inputs and maps them to target space so that the $L_2$ norm distance in target space can represent the semantic distance in input space. The mapping is represented by the neural network in supervised learning. Moreover, a more lightweight Siamese network model, SiameseFace2, is designed to reduce the network parameters without losing accuracy. We also present a new method to generate training data and expand the number of training samples for single category in AR and labeled faces in the wild (LFW) datasets, which improves the recognition accuracy of the models. Four loss functions are adopted to carry out experiments on AR and LFW datasets. The results show that the contrastive loss function combined with new Siamese network model in this paper can effectively improve the accuracy of face recognition.

• 한국지리정보학회지
• /
• 제1권1호
• /
• pp.3-7
• /
• 1998

GIS 데이타의 정확도 문제는 DEM과 같은 데이터의 유용성과 적용에 대한 서로 다른 견해를 불러 일으킨다. 데이터의 정확성은 좌표의 정확한 위치와 속성정보를 무작위적으로 검색하여 결정할 수 있다. DEM은 과거 보다는 손쉽게 취득할수 있고 이를 처리할수 있는 소프트웨어도 다양해 졌으나 GIS의 응용은 이미 만들어진 데이터에 따라 그 결과가 달라질수 있으므로 데이터의 정확도와 에러에 대한 주의를 기울일 필요가 있다. 본 연구의 목적은 1:24,000과 1:250,000 DEM 데이터를 이용하여 DEM의 정확도를 검색하고 데이터가 지닌 에러를 찾아내는 방법을 모색하는데 있다. GRASS와 Arc/Info를 이용하여 DEM을 레이어로 만들어내는 과정 또한 연구 되었다. 연구지역은 250 $km^2$의 면적을 지녔으며 연구 결과 1:250,000 DEM에서는 실제 등고값이 정상적으로 처리 되었으나 1:24,000 DEM에서는 실제의 등고값이 아닌 0으로 표현된 에러가 발견되었다.

• 대한치과의사협회지
• /
• 제57권12호
• /
• pp.747-756
• /
• 2019

This paper reviews the adaptation accuracy and mechanical properties of currently used denture processing systems with base resin materials and introduces the latest research on the development of antimicrobial denture base resins. Poly(methyl methacrylate) has been successfully used as a dental denture base resin material by the compress-molding method and heat polymerization for a long time, but recently, new processing techniques, injection molding-methods or fluid-resin technique are also used for fabricating denture base. However, studies indicated that there was no difference between the injectionmolding and the conventional compression-molding method in terms of adaption accuracy of denture base. The fluid-resin fabrication and one injection-molding systems exhibited better adaptation accuracy than the other processing methods. Resin denture bases in the oral cavity may undergo midline fractures due to flexural fatigue from repeated masticatory loading. For those patients, impact resistant denture base resins are recommended to prevent denture fracture during service. Thermoplastic denture base resins can be helpful for patients suffering from allergic reaction to resin monomers with a soft-fit, however, thermoplastic resins with low stiffness can irritate gum tissues and accelerate abnormal alveolar ridge resorption. Moreover, due to low chemical durability in oral cavity, those should be used for a limited period of time.

• The Journal of Advanced Prosthodontics
• /
• 제13권2호
• /
• pp.89-99
• /
• 2021

PURPOSE. This study aimed to compare the accuracy (trueness and precision) of interim crowns fabricated using DLP (digital light processing) according to post-curing time. MATERIALS AND METHODS. A virtual stone study die of the upper right first molar was created using a dental laboratory scanner. After designing interim crowns on the virtual study die and saving them as Standard Triangulated Language files, 30 interim crowns were fabricated using a DLP-type 3D printer. Additively manufactured interim crowns were post-cured using three different time conditions-10-minute post-curing interim crown (10-MPCI), 20-minute post-curing interim crown (20-MPCI), and 30-minute post-curing interim crown (30-MPCI) (n = 10 per group). The scan data of the external and intaglio surfaces were overlapped with reference crown data, and trueness was measured using the best-fit alignment method. In the external and intaglio surface groups (n = 45 per group), precision was measured using a combination formula exclusive to scan data (10C2). Significant differences in accuracy (trueness and precision) data were analyzed using the Kruskal-Wallis H test, and post hoc analysis was performed using the Mann-Whitney U test with Bonferroni correction (α=.05). RESULTS. In the 10-MPCI, 20-MPCI, and 30-MPCI groups, there was a statistically significant difference in the accuracy of the external and intaglio surfaces (P<.05). On the external and intaglio surfaces, the root mean square (RMS) values of trueness and precision were the lowest in the 10-MPCI group. CONCLUSION. Interim crowns with 10-minute post-curing showed high accuracy.

• 대한치과기공학회지
• /
• 제44권2호
• /
• pp.31-37
• /
• 2022

Purpose: The aim of this study was to evaluate the accuracy of provisional crowns manufactured using a milling machine and a digital light processing (DLP) printer. Methods: A full-contour crown was designed using computer-aided design software. Provisional crowns of this design were manufactured using a milling machine and using a DLP three-dimensional (3D) printer (N=20). The provisional crowns were digitized with an extraoral scanner, and 3D deviation analysis was applied to the scanned data to confirm their accuracy. An independent t-test was performed to detect the significant differences, and the Kolmogorov-Smirnov test was used for analysis (α=0.05). Results: No significant differences were found among the precision of marginal surface between the printed and milled crowns (p=0.181). The trueness of marginal and internal surfaces of the milled crowns were statistically higher than those of the printed crowns (p=0.024, p=0.001; respectively). Conclusion: The accuracy of provisional crowns manufactured using a milling machine and a 3D printer differed significantly except with regards to the precision of the internal surface. However, all the crowns were clinically acceptable, regardless of the manufacturing method used.

• 대한기계학회논문집
• /
• 제18권5호
• /
• pp.1182-1191
• /
• 1994

This paper presents an enhanced method of slit beam projection intended for the rapid measurement of 3-dimensional surface profiles of dies and molds. Special emphasis is given to optimizing the design of optical system so that the measuring accuracy can be maximized by adopting two-plane camera calibration together with sub-pixel image processing techniques. Finally, several measurement examples are discussed to demonstrate that an actual measuring accuracy of $\pm$ 0.2 mm can be achieved over the measuring range of 500 mm{\times}300mm{\times}200mm$.

• 한국전기전자재료학회논문지
• /
• 제19권4호
• /
• pp.386-390
• /
• 2006

Global consumption of aspheric lens will expand rapidly due to golbal transformation of the electronics based industry to optics based mechatronics. Especially, F-Theta lens is one of important parts in Laser Scanning Unit(LSU) because it affects the optical performance of LSU dominantly. Non axisymmetric machine based processing techologies are required to obtain high accuracy in utlra-precision aspheric core, the most important component in plastic injection molded F-Theta lens assembly. In this study, the core with non-axisymmetric aspheric shape which is used to emit the F-Theta lens was processed using the ultra precision processing technology and the shape accuracy of the core was measured. And the results there of were evaluated and compared with the emitted shape accuracy of F-Theta lens.

• 한국정보처리학회논문지
• /
• 제7권4호
• /
• pp.1150-1159
• /
• 2000

This paper presents an automatic text categorization model that improves the accuracy by combining statistical and knowledge-based categorization methods. In our model we apply knowledge-based method first, and then apply statistical method on the text which are not categorized by knowledge-based method. By using this combined method, we can improve the accuracy of categorization while categorize all the texts without failure. For statistical categorization, the vector model with Inverted Category Frequency (ICF) weighting is used. For knowledge-based categorization, Phrasal Patterns and Keyword Sets are introduced to represent sentence patterns, and then pattern matching is performed. Experimental results on new articles show that the accuracy of categorization can be improved by combining the tow different categorization methods.