• 제목/요약/키워드: time averaging

검색결과 337건 처리시간 0.024초

패턴인식을 이용한 수삼 등급판정 알고리즘에 관한 연구 (A Study on a Ginseng Grade Decision Making Algorithm Using a Pattern Recognition Method)

  • 정석훈;고국원;강제용;장수원;이상준
    • 정보처리학회논문지:소프트웨어 및 데이터공학
    • /
    • 제5권7호
    • /
    • pp.327-332
    • /
    • 2016
  • 본 연구는 비 정형 농산물 중 6년근 수삼의 자동 등급 분류하기 위한 선행연구로, 이를 위해 4방향에서 이미지 취득이 가능한 수삼 영상 측정기를 제작 하였으며 총 245 수삼 개체에 대해서 영상을 취득하였다. 취득된 영상의 각 수삼 개체마다 12개의 파라미터를 추출하였으며, KGC 인삼공사의 수삼등급 분류 기준과 각 등급별 평균 파라미터의 분포를 조사하여 최종 4개 파라미터를 선정하였다. 패턴인식 분류기는 Support Vector Machine을 사용하였으며 공용 소프트웨어인 OpenCV Library를 사용하여 k-Class 분류기를 설계하였다. 각 등급별 학습 데이터 수를 10, 15, 20으로 조정하여 등급별 인식률, 본인 거부율, 타인 인식율을 조사하였으며, 학습데이터 수가 10개일 때 1등급 인식률 94%, 2등급 인식률 98%, 3등급 인식률 90%로 가장 높은 인식 성능을 보였다.

반능동 단속형 감쇠기를 이용한 현가장치 개선에 관한 연구 (A study on the improvement of a suspension system adopting a semiactive on-off damper)

  • 최성배;박윤식
    • 대한기계학회논문집
    • /
    • 제12권5호
    • /
    • pp.959-967
    • /
    • 1988
  • 본 연구에서는 반능동형 감쇠기의 작동기준을 설정하고 그것의 타당성을 조사 하는 것이 주목적이어서 속도에 대한 고려는 제외하였다.이때 반능동형 감쇠계가 고정된 감쇠계수를 갖는 감쇠기로 구성된 계보다 얼마만큼 성능이 향상되며 능동형 감 쇠기(감쇠계수를 제한된 영역내에서 순간순간 조절하여 변화시킬 수 있는 감쇠기)를 갖는 계에 얼마만큼 접근하는가가 비교되어진다.

생물학적 수소생산을 위한 Trickling Bed Biofilter에서의 친수성과 소수성 담체의 영향 (Effect of Hydrophilic- and Hydrophobic-Media on the Fermentative Hydrogen Production in Trickling Bed Biofilter)

  • 전병승;이선미;김용환;채희정;상병인
    • 한국신재생에너지학회:학술대회논문집
    • /
    • 한국신재생에너지학회 2006년도 춘계학술대회
    • /
    • pp.465-469
    • /
    • 2006
  • Two mesophilic trickling bed bioreactors filled with two different types of media, hydrophilic- and hydrophobic-cubes, were designed and tested for hydrogen production via anaerobic fermentation of sucrose. Each reactor consisted of a column packed with polymeric cubes and inoculated with heat-treated sludge obtained from anaerobic digestion tank. A defined medium containing sucrose was fed with changing flow rate into the capped reactor, hydraulic retention time and recycle rate. Hydrogen concentrations in gas-phase were constant, averaging 40% for all conditions tested. Hydrogen production rates increased up to $10.5 L{\cdot};h^{-1}{\cdot}L^{-1}$ of reactor when influent sucrose concentrations and recycle rates were varied. Hydrophobic media provided higher value of hydrogen production rate than hydrophilic media at the same operation conditions. No methane was detected when the reactor was under a normal operation. The major fermentation by-products in the liquid effluent of the both trickling biofilters were acetate and butyrate. The reactor filled with hydrophilic media became clogged with biomass and bio gas, requiring manual cleaning of the system, while no clogging occurred in the reactor with hydrophobic media. In order to make long-term operation of the reactor filled with hydrophilic media feasible, biofilm accumulation inside the media in the reactor with hydrophilic media and biogas produced from the reactor will need to be controlled through some process such as periodical backwashing or gas-purging. These tests using trickling bed biofilter with hydrophobic media demonstrate the feasibility of the process to produce hydrogen gas in a trickle-bed type of reactor. A likely application of this reactor technology could be hydrogen gas recovery from pre-treatment of high carbohydrate-containing wastewaters.

  • PDF

뇌동맥류 파열 환자의 수술후 인지기능과 기억력장애에 관한 연구 (Cognition and Memory Impairment after Operation in Ruptured Cerebral Aneurysm Patients)

  • 김병주;최창화;김대진
    • Journal of Korean Neurosurgical Society
    • /
    • 제30권7호
    • /
    • pp.842-848
    • /
    • 2001
  • Objectives : The mortality rate of subarachnoid hemorrhage(SAH) has been reduced recently due to refinement of microsurgical technique and improved perioperative management. Also, many survivors of SAH show excellent neurological recoveries. However, we found that a high proportion of the survivors do not fully regain their premorbid status in cognitive and memory function. Object of this study is to evaluate which factors might influence on cognitive and memory impairment in ruptured aneurysmal SAH patients. Methods : In this prospective study, a series of 66 patients with aneurysmal subarachnoid hemorrhage(SAH) from 1996 to 1998, most of whom had a "good" or "fair" neurological outcome, were assessed with various tests of cognition and memory function. All patients underwent clipping operation by pterional approach. Right side approach was performed in 16 case and left 21 cases. K-WAIS(Korean-Wechsler Adult Intelligence Scale) was used as method of cognition and memory function test. The time interval between SAH and assessment varied between 4 months and 8 months, averaging 6.2 months. Statistical analyses were carried out for each test score to see whether aneurysm site(A-com : non A-com), route of approach, age and sex, vasospasm, Hunt-Hess grade and Fisher CT group at admission, Glasgow Outcome Scale(GOS) at discharge affect cognitive and memory function. Results : Aneurysm site was not shown to be associated with performance on any test, and the initial grade (Hunt-Hess grade, Fisher CT group) of SAH and vasospasm had only minimal predictive values. The grade at discharge( GOS) was proved to be the best predictor of impairment of cognition and memory function within 1 year after operation. Conclusion : The authors conclude that the diffuse effects of SAH are more important than focal neuropathology in relation to cognitive impairment in this group of patients.

  • PDF

저화질 영상 인식을 위한 화질 저하 모델 기반 다중 인식기 결합 (Multiple-Classifier Combination based on Image Degradation Model for Low-Quality Image Recognition)

  • 류상진;김인중
    • 정보처리학회논문지B
    • /
    • 제17B권3호
    • /
    • pp.233-238
    • /
    • 2010
  • 본 논문에서는 화질 저하 모델에 기반한 다중 인식기 결합을 이용하여 저화질 영상에 대한 인식 성능을 개선하기 위한 방법을 제안한다. 제안하는 방법은 화질 저하 모델을 이용해 특정 화질에 각각 특화된 복수의 인식기들을 생성한다. 인식 과정에서는 인식기들의 결과를 가중 평균에 의해 결합함으로써 최종 결과를 결정한다. 이 때, 각 인식기의 가중치는 입력 영상의 화질 추정 결과에 따라 동적으로 결정된다. 입력 영상의 화질에 특화된 인식기에는 큰 가중치를, 그렇지 않은 인식기에는 작은 가중치를 지정한다. 그 결과, 입력 영상의 화질 변이에 효과적으로 적응할 수 있다. 뿐만 아니라, 복수의 인식기를 사용하기 때문에 저화질 영상에 대하여 단일 인식 시스템보다 더욱 안정적인 성능을 나타낸다. 제안하는 다중 인식기 결합 방법은 화질을 고려하지 않은 다중 인식기 결합 방법이나, 화질을 고려한 단일 인식 방법과 비교하여 더 높은 인식률을 보였다.

Denoise of Astronomical Images with Deep Learning

  • Park, Youngjun;Choi, Yun-Young;Moon, Yong-Jae;Park, Eunsu;Lim, Beomdu;Kim, Taeyoung
    • 천문학회보
    • /
    • 제44권1호
    • /
    • pp.54.2-54.2
    • /
    • 2019
  • Removing noise which occurs inevitably when taking image data has been a big concern. There is a way to raise signal-to-noise ratio and it is regarded as the only way, image stacking. Image stacking is averaging or just adding all pixel values of multiple pictures taken of a specific area. Its performance and reliability are unquestioned, but its weaknesses are also evident. Object with fast proper motion can be vanished, and most of all, it takes too long time. So if we can handle single shot image well and achieve similar performance, we can overcome those weaknesses. Recent developments in deep learning have enabled things that were not possible with former algorithm-based programming. One of the things is generating data with more information from data with less information. As a part of that, we reproduced stacked image from single shot image using a kind of deep learning, conditional generative adversarial network (cGAN). r-band camcol2 south data were used from SDSS Stripe 82 data. From all fields, image data which is stacked with only 22 individual images and, as a pair of stacked image, single pass data which were included in all stacked image were used. All used fields are cut in $128{\times}128$ pixel size, so total number of image is 17930. 14234 pairs of all images were used for training cGAN and 3696 pairs were used for verify the result. As a result, RMS error of pixel values between generated data from the best condition and target data were $7.67{\times}10^{-4}$ compared to original input data, $1.24{\times}10^{-3}$. We also applied to a few test galaxy images and generated images were similar to stacked images qualitatively compared to other de-noising methods. In addition, with photometry, The number count of stacked-cGAN matched sources is larger than that of single pass-stacked one, especially for fainter objects. Also, magnitude completeness became better in fainter objects. With this work, it is possible to observe reliably 1 magnitude fainter object.

  • PDF

Comparison of aerodynamic loading of a high-rise building subjected to boundary layer and tornadic winds

  • Ashrafi, Arash;Chowdhury, Jubayer;Hangan, Horia
    • Wind and Structures
    • /
    • 제34권5호
    • /
    • pp.395-405
    • /
    • 2022
  • Tornado-induced damages to high-rise buildings and low-rise buildings are quite different in nature. Tornado losses to high-rise buildings are generally associated with building envelope failures while tornado-induced damages to low-rise buildings are usually associated with structural or large component failures such as complete collapses, or roofs being torn off. While studies of tornado-induced structural damages tend to focus mainly on low-rise residential buildings, transmission towers, or nuclear power plants, the current rapid expansion of city centers and development of large-scale building complexes increases the risk of tornadoes impacting tall buildings. It is, therefore, important to determine how tornado-induced load affects tall buildings compared with those based on synoptic boundary layer winds. The present study applies an experimentally simulated tornado wind field to the Commonwealth Advisory Aeronautical Research Council (CAARC) building and estimates and compares its pressure coefficient effects against the Atmospheric Boundary Layer (ABL) flow field. Simulations are performed at the Wind Engineering, Energy and Environment (WindEEE) Dome which is capable of generating both ABL and tornadic winds. A model of the CAARC building at a scale of 1:200 for both ABL and tornado flows was built and equipped with pressure taps. Mean and peak surface pressures for TLV flow are reported and compared with the ABL induced wind for different time-averaging. By following a compatible definition of the pressure coefficients for TLV and ABL fields, the resulting TLV pressure field presents a similar trend to the ABL case. Also, the results show that, for the high-rise building model, the mean and 3-sec peak pressures are larger for the ABL case compared to the TLV case. These results provide a way forward for the code implementation of tornado-induced pressures on high-rise buildings.

한국어판 난임 낙인 측정도구의 타당도와 신뢰도 (The Reliability and Validity of Korean Version of the Infertility Stigma Scale (K-ISS))

  • 김미옥;반민경
    • 대한간호학회지
    • /
    • 제52권6호
    • /
    • pp.582-597
    • /
    • 2022
  • Purpose: This study aimed to translate the Infertility Stigma Scale (ISS) into Korean and to evaluate its reliability and validity in the Korean context. Methods: Data were collected from 350 women who underwent in vitro fertilization (IVF). Data were analyzed using SPSS WIN 25.0 and AMOS 22.0. Content validity was analyzed using the item-level content validity index (I-CVI) and scale-level content validity index/averaging (S-CVI/Ave). The preliminary survey was conducted on 20 women who had experienced IVF at least once to check the level of understanding of the tool and the time required to fill out the questionnaire. Exploratory and confirmatory factor analyses were used to test construct validity. Additionally, hypothesis-testing construct validity were tested. Cronbach's α was used to assess the reliability. Results: The Korean-ISS (K-ISS) consists of 25 items, excluding two items from the original ISS questionnaire. Exploratory factor analysis identified four factors, which explained 75.6% of the total variance. The four distinct factors were infertility stigma with self-devaluation (56.8%), public stigma (8.1%), social withdrawal (6.5%), and family stigma (4.2%). In the confirmatory factor analysis, the 25 items in the four-factor structure were validated (χ2 /df ≤ 3, RMSEA ≤ 10). The hypothesis-testing construct validity of K-ISS against FPI (r = .58~.71, p < .001) and FQI (r = - .49~- .65, p < .001) was tested and found to be significant. The internal consistency reliability of the K-ISS, assessed using Cronbach's α, was .97. Conclusion: The K-ISS has satisfactory construct validity and reliability; therefore, it can help minimize the negative impact of stigma by measuring the stigma associated with women experiencing infertility.

트랜스포머 기반 효율적인 자연어 처리 방안 연구 (A Study on Efficient Natural Language Processing Method based on Transformer)

  • 임승철;윤성구
    • 한국인터넷방송통신학회논문지
    • /
    • 제23권4호
    • /
    • pp.115-119
    • /
    • 2023
  • 현재의 인공지능에서 사용되는 자연어 처리 모델은 거대하여 실시간으로 데이터를 처리하고 분석하는 것은 여러가지 어려움들을 야기하고 있다. 이런 어려움을 해결하기 위한 방법으로 메모리를 적게 사용해 처리의 효율성을 개선하는 방법을 제안하고 제안된 모델의 성능을 확인하였다. 본 논문에서 제안한 모델의 성능평가를 위해 적용한 기법은 BERT[1] 모델의 어텐션 헤드 개수와 임베딩 크기를 작게 조절해 큰 말뭉치를 나눠서 분할 처리 후 출력값의 평균을 통해 결과를 산출하였다. 이 과정에서 입력 데이터의 다양성을 주기위해 매 에폭마다 임의의 오프셋을 문장에 부여하였다. 그리고 모델을 분류가 가능하도록 미세 조정하였다. 말뭉치를 분할 처리한 모델은 그렇지 않은 모델 대비 정확도가 12% 정도 낮았으나, 모델의 파라미터 개수는 56% 정도 절감되는 것을 확인하였다.

Apparent digestibility coefficients of animal feed ingredients for olive flounder (Paralichthys olivaceus)

  • Md Mostafizur Rahman;Kang-Woong Kim;Sang-Min Lee
    • Fisheries and Aquatic Sciences
    • /
    • 제25권11호
    • /
    • pp.537-548
    • /
    • 2022
  • Apparent digestibility coefficients (ADCs) of dry matter, crude protein, crude lipid, nitrogen-free extract, energy and essential amino acids in animal-based feed ingredients were determined for olive flounder (Paralichthys olivaceus). A reference diet (RF) was formulated to contain 1.0% chromic oxide (Cr2O3) as an inert indicator. Nine test diets were formulated to contain RF and one of the feed ingredients (pollock meal [PM], jack mackerel meal [JMM], anchovy meal [AM], cod meal [CM], sardine meal [SM], sand eel meal [SEM], tuna meal [TM], meat meal [MM] and squid liver meal [SLM]) at a 7:3 ratio in each diet designated as PM, JMM, AM, CM, SM, SEM, TM, MM and SLM, respectively. Olive flounder, averaging 150 ± 8.0 g, were stocked at a density of 25 fish per tank in 400-L fiberglass tanks attached with fecal collection columns. Feces were collected from triplicate groups of fish one time a day for four weeks. Dry matter and crude protein ADCs of CM and SEM were significantly higher than the other tested ingredients. Lipid ADCs of JMM, CM and SEM were significantly higher than the other test ingredients. Energy ADCs of CM and SEM were significantly higher than that of the other tested ingredients. The availability of amino acids in CM was generally higher than the other animal protein sources. PM exhibited the lowest amino acid availability among the treatments. Interestingly, MM exhibited significantly higher nutrient digestibility than several marine-based ingredients. However, CM and SEM are seeming to be highly digestible and effective to use in olive flounder diet compared to the other tested ingredients. Overall, the results of this study provide information about the bioavailability of nutrients and energy in animal feedstuffs to apply when formulating cost-effective practical feeds for olive flounder.