• 한국컴퓨터정보학회논문지
• /
• 제19권11호
• /
• pp.43-52
• /
• 2014

본 논문에서는 적외선 영상에서 영상 변위를 이용하여 기동 표적 영역을 탐지하고, SURF(Speeded Up Robust Features) 특징점에 대한 BAS(Beam Angle Statistics)를 이용하여 분류하는 시스템에 대하여 설명한다. 영상 기반 기술 분야에서 대표적인 대응점 정합 알고리즘인 SURF 기법은 SIFT(Scale Invariant Feature Transform) 기법에 비해 정합 속도가 매우 빠르고 비슷한 정합 성능을 보이기 때문에 널리 사용되고 있다. SURF를 이용한 대부분의 객체 인식의 경우 특징점 추출과 정합의 과정을 수행하지만, 제안하는 기법은 표적의 기동 특성을 반영하여 영상의 변위 추정을 통하여 표적의 영역을 탐지하고 SURF 특징점 들의 기하구조를 판단함으로써 표적 분류를 수행한다. 제안하는 기법은 무인 표적 탐지/인지 시스템의 초기모델 구축을 위하여 연구가 진행되었으며, 모의 표적을 이용한 가상 영상과 적외선 실 영상을 이용하여 실험한 결과 약 73~85%의 분류 성능을 확인하였다.

• 약학회지
• /
• 제41권4호
• /
• pp.421-432
• /
• 1997

A new lectin was partially purified from starfish,Asterina pectinifera by means of physiological saline extraction, salt fractionation, ion exchange chromatography and hy droxyapatite chromatography, and it was named APL. The biochemical properties of the APL were characterized. In addition, its effects on lymphocyte mitogenicity and cancer cell agglutinability were tested. The APL agglutinated nonspecifically human erythrocytes and rabbit blood cells. Agglutinability was decreased to 30% of control activity below pH 5 and above pH 9 and was relatively unstable at increasing temperatures above 60$^{\circ}C$. The activity was reduced by addition of two kinds of metal ions, $Ba^{2+},\;Mn^{2+}$ and chelating agent, EDTA. APL was proved to be glycoproteins containing 9% sugars. For carbohydrate specificity, it was found that the activity of APL was inhibited by D(+)-glucosamine, D(+)-galactosamine, stachyose, N-acetyl-galactosamine and methyl-${\alpha}$-D-galactopyranoside among 35 sugars tested. In amino acid composition, the contents of acidic amino acids such as aspartic acid and glutamic acid were relatively high. This result suggest that the isoelectric point would be in a lower range. APL was found that it promotes the division of human lymphocytes. APL was proved to be a potent agglutinin for cancer cells such as HeLa, L929 and L1210 cells. Significant changes on the HeLa cell surfaces affected by APL were observed under the electron microscope.

• 대한치과교정학회지
• /
• 제51권5호
• /
• pp.313-320
• /
• 2021

Objective: The purpose of this study was to evaluate the effects of chewing gum and low-level laser therapy in alleviating orthodontic pain induced by the initial archwire. Methods: Patients with 3-6 mm maxillary crowding who planned to receive non-extraction orthodontic treatment were recruited for the study. Sixty-three participants (33 females and 30 males) were randomly allocated into three groups: laser, chewing gum, and control. In the laser group, a gallium aluminum arsenide (GaAlAs) diode laser with a wavelength of 820 nm was used to apply a single dose immediately after orthodontic treatment began. In the chewing gum group, sugar-free gum was chewed three times for 20 minutes-immediately after starting treatment, and at the twenty-fourth and forty-eighth hours of treatment. Pain perception was measured using a visual analog scale at the second, sixth, and twenty-fourth hours, and on the second, third, and seventh days. Results: There were no statistically significant differences between the groups at any measured time point (p > 0.05). The highest pain scores were detected at the twenty-fourth hour of treatment in all groups. Conclusions: Within the limitations of the study, we could not detect whether low-level laser therapy and chewing gum had any clinically significant effect on orthodontic pain. Different results may be obtained with a higher number of participants or using lasers with different wavelengths and specifications. Although the study had a sufficient number of participants according to statistical analysis, higher number of participants could have provided more definitive outcomes.

• 산업경영시스템학회지
• /
• 제44권3호
• /
• pp.98-105
• /
• 2021

The sensory stimulation of a cosmetic product has been deemed to be an ancillary aspect until a decade ago. That point of view has drastically changed on different levels in just a decade. Nowadays cosmetic formulators should unavoidably meet the needs of consumers who want sensory satisfaction, although they do not have much time for new product development. The selection of new products from candidate products largely depend on the panel of human sensory experts. As new product development cycle time decreases, the formulators wanted to find systematic tools that are required to filter candidate products into a short list. Traditional statistical analysis on most physical property tests for the products including tribology tests and rheology tests, do not give any sound foundation for filtering candidate products. In this paper, we suggest a deep learning-based analysis method to identify hand cream products by raw electric signals from tribological sliding test. We compare the result of the deep learning-based method using raw data as input with the results of several machine learning-based analysis methods using manually extracted features as input. Among them, ResNet that is a deep learning model proved to be the best method to identify hand cream used in the test. According to our search in the scientific reported papers, this is the first attempt for predicting test cosmetic product with only raw time-series friction data without any manual feature extraction. Automatic product identification capability without manually extracted features can be used to narrow down the list of the newly developed candidate products.

• 한국인터넷방송통신학회논문지
• /
• 제19권1호
• /
• pp.145-152
• /
• 2019

본 논문에서는 딥러닝을 활용한 흔들림 영상 안정화 알고리즘을 제안하였다. 제안하는 알고리즘은 기존 몇 가지 2D, 2.5D 및 3D 기반 안정화 기술과 다르게 딥러닝을 활용한다. 제안하는 알고리즘은 흔들리는 영상을 CNN 네트워크 구조와 LSTM 네트워크 구조를 통한 특징 추출 및 비교하여 이전 프레임과 현재 프레임 간의 특징점 위치 차이를 통해 특징점의 이동 크기와 방향의 반대로 영상을 변환하는 알고리즘이다. 흔들림 안정화를 위한 알고리즘은 각 프레임의 특징 추출 및 비교를 위해 Tensorflow를 활용하여 CNN 네트워크과 LSTM 구조를 구현하였으며, 영상 흔들림 안정화는 OpenCV open source를 활용해 구현하였다. 실험결과 영상의 흔들림이 상하좌우로 흔들리는 영상과, 급격한 카메라 이동이 없는 영상을 실험에 사용하여, 제안한 알고리즘을 적용한 결과 사용한 상하좌우 흔들림 영상에서는 안정적인 흔들림 안정화 성능을 기대할 수 있었다.

• 한국환경농학회지
• /
• 제39권4호
• /
• pp.368-374
• /
• 2020

BACKGROUND: The residual analysis of polar pesticides has remained a challenge. It is even more difficult to simultaneously analyze multiple polar pesticides. Diquat, paraquat, and chlormequat are typical examples of highly polar pesticides. The existing methods for the analysis of diquat, paraquat and chlormequat are complex and time consuming. Therefore, a simple, quick and effective method was developed in the represent study for simultaneous analysis of diquat, paraquat and chlormequat in animal products, meat and fat using UPLC-MS/MS. METHODS AND RESULTS: Sample extraction was carried out using acidified acetonitrile and water and re- extracted with acidified acetonitrile and combine the extracts followed by centrifugation. The extract was then cleaned up with a HLB cartridge after reconstitution with acidic acetonitrile and water. The method was validated in quintuplicate at three different concentrations. The limits of detection (LOD) and quantification (LOQ) were 0.0015 and 0.005 mg/L, respectively. Matrix suppression effect was observed for all of the analytes. A seven point matrix matched calibration curve was constructed for each of the compound resulted excellent linearity with determination coefficients (R2) ≥ 0.991. Accuracy and precision of the method was calculated from the recovery and repeatability and ranged from 62.4 to 119.7% with relative standard deviation less than 18.8%. CONCLUSION: The recovery and repeatability of the developed method were in the acceptable range according to the Codex Alimentarius guideline. The developed method can be applied for the routine monitoring of diquat, paraquat, and chlormequat in animal products, meat and fat.

• 한국측량학회지
• /
• 제38권6호
• /
• pp.553-559
• /
• 2020

노후화된 지하시설물의 손상이 주요 원인이 되어 발생하는 도심지 지반침하를 예방하기 위해 지하시설물의 체계적 관리를 위해 지하공간 통합지도가 제작되고 있다. 그러나, 지하공간 통합지도를 갱신하는 과정에 갱신되지 않은 지하시설물 객체가 포함되어 갱신 시간을 지연시키는 문제가 있다. 본 논문에서는 지하시설물의 중심점을 기반으로 지하공간 통합지도의 갱신 과정에 필요한 갱신된 객체만을 선별함으로써 통합지도의 갱신 시간을 단축하는 방법에 대해 제안하였다. 중심점 비교를 통해 검색 대상을 대폭 감소시켜 검색 속도를 단축하고, 동일 객체 여부는 두 객체의 실제 위치를 이용한 거리를 계산하여 판정하였다. 제안된 방법은 데이터의 수량이 많을수록 빠른 성능을 나타내고, 기존의 방법에 비해 약 4배 정도 빠르게 갱신된 객체를 지하공간 통합지도에 반영할 수 있다.

• 한국정보통신학회논문지
• /
• 제23권12호
• /
• pp.1528-1534
• /
• 2019

최근 들어, 3차원의 깊이 카메라의 대중화로 인해서 RGB 영상에서 수행되던 연구에 새로운 관심과 기회가 생겼지만 사람의 손 자세의 추정은 여전히 어려운 주제 중의 하나로 분류되고 있다. 본 논문에서는 다양하게 입력되는 3차원의 깊이 영상으로부터 사람의 손의 자세를 학습 알고리즘을 이용하여 강인하게 추정하는 방법을 제안한다. 제안된 접근 방법에서는 먼저 뼈대 기반의 손 모델을 생성한 다음, 생성된 손 모델을 3차원의 포인트 클라우드 데이터에 정렬한다. 그런 다음, 랜덤 포레스트 기반의 학습 알고리즘을 이용하여 정렬된 손 모델로부터 손의 자세를 강인하게 추정한다. 본 논문의 실험 결과에서는 제안된 접근 방법이 다양한 실내외의 환경에서 촬영된 입력 영상으로부터 사람의 손의 자세를 강인하고 빠르게 추정한다는 것을 보여준다.

• 한국ITS학회 논문지
• /
• 제20권6호
• /
• pp.47-54
• /
• 2021

국내 고속도로 교통사고 건수는 2020년 기준 약 4천건으로, 비반복적 정체와 높은 주행속도로 인해 다른 도로 대비 교통사고 발생 건수 당 사망자 수는 약3.7배에 달한다. 고속도로의 사고 유형은 측면충돌 및 추돌사고가 대부분을 차지하며, 주요 요인 중 하나는 분·합류부, 사고 등으로 야기되는 위험 교통류라고 할 수 있다. 따라서, 고속도로와 같은 연속류에서 나타나는 위험 교통류는 운전자에게 사고 방지를 위한 중요한 정보라고 할 수 있다. 본 연구에서는 개별차량 정보를 이용하여 속도의 변화 지점과 차로별 속도 차이가 발생하는 구간 등 위험 교통류를 분류하고자 하였다. 지오해시 기반으로 공간을 분리하였으며, 동일 구간 내에서 개별 차량의 속도 차이를 나타낼 수 있는 공간평균속도와 차량간 속도 편차를 이용하여 속도의 동질 구간을 분류하였다. 그 결과 고속도로 위험 구간 정보를 제공할 수 있는 분류부 영향권 구간과 위험 교통류 구간을 추출하였다.

• Smart Structures and Systems
• /
• 제30권6호
• /
• pp.687-701
• /
• 2022

Inevitable response loss under complex operational conditions significantly affects the integrity and quality of measured data, leading the structural health monitoring (SHM) ineffective. To remedy the impact of data loss, a common way is to transfer the recorded response of available measure point to where the data loss occurred by establishing the response mapping from measured data. However, the current research has yet addressed the structural condition changes afterward and response mapping learning from a small sample. So, this paper proposes a novel data driven structural response reconstruction method based on a sophisticated designed generating adversarial network (UAGAN). Advanced deep learning techniques including U-shaped dense blocks, self-attention and a customized loss function are specialized and embedded in UAGAN to improve the universal and representative features extraction and generalized responses mapping establishment. In numerical validation, UAGAN efficiently and accurately captures the distinguished features of structural response from only 40 training samples of the intact structure. Besides, the established response mapping is universal, which effectively reconstructs responses of the structure suffered up to 10% random stiffness reduction or structural damage. In the experimental validation, UAGAN is trained with ambient response and applied to reconstruct response measured under earthquake. The reconstruction losses of response in the time and frequency domains reached 16% and 17%, that is better than the previous research, demonstrating the leading performance of the sophisticated designed network. In addition, the identified modal parameters from reconstructed and the corresponding true responses are highly consistent indicates that the proposed UAGAN is very potential to be applied to practical civil engineering.