교통안전시설물의 관리는 도로교통의 안전과 직결되는 문제이다. 운전자는 신호등, 표지판, 노면표시 등을 통해 운전에 필요한 정보 를 얻는다. 노후된 표지판과 노면표시는 운전자에게 잘못된 정보를 제공할 수 있으므로 주기적인 시설물의 관리가 필요하다. 본 연구 는 딥 러닝 기술을 활용해 운전자 시각의 영상 자료에서 교통안전표지를 자동으로 탐지하고자 하며, 교통안전표지의 공통된 색상 특 징을 기반으로 클래스를 그룹으로 묶어 데이터셋을 구축하는 방법을 제안한다. 객체탐지의 성능지표로 널리 활용되는 mAP를 사용해 클래스 묶음 여부에 따른 탐지 성능을 비교한 결과, 색상 기반 클래스 묶음을 적용한 모델의 탐지 성능이 비교군에 비해 약 36% 상승 함을 확인하였다.

도로포장의 대표적 파손 종류인 균열은 일반적으로 폭이 좁고 기하학적으로 정의하기 어렵기 때문에 균열을 검출하고 유형을 분류 한 후 정량화하기까지 많은 시간이 소요된다. 본 연구의 목적은 균열 검출 이후 단계에서 요구되는 분류 및 정량화 과정을 자동화하 기 위함이다. 이를 위해, 본 연구에서는 균열이 매핑된 포장관리체계용 노면영상을 대상으로 하는 25cm 정사각형의 격자 배치 방법과 차륜 통과 영역 구분을 제시하였다. 각 격자 내 균열 객체의 길이와 진전방향, 인접한 정도 등 시각적 정보에 의한 균열 격자 속성을 정의하고 프로그래밍하여 균열 유형분류와 집계를 자동화하였다. 무작위로 수집된 고속도로 노면영상 자료를 통해 포장형식 별 주요 균열 유형을 분석하였고 차륜 통과 영역에서의 균열률 증가를 수치적으로 확인하였다.

포장상태 평가를 위한 노면영상 촬영은 라인스캔 방식이 주를 이루고 있다. 라인스캔 특성 상, 조사환경이나 장비특성이 달라질 경 우 밝기가 상이한 노면영상을 취득할 수 있고 이는 U-net과 같은 픽셀 단위 segmentation 딥러닝 모델의 균열 자동검출 성능에 영향을 미친다. 본 연구에서는 인공지능 검출 모델의 변경 없이 영상의 밝기 최적화와 morphology 연산기법을 노면영상 전·후처리 방법으로 제시하고 그 효과를 분석하였다. 영상 처리를 통해 과다 검출경향을 보인 이상치들이 제거되었으며 정답으로 간주할 수 있는 전문요 원 분석결과인 GT 균열률과의 상관성 또한 향상됨을 확인하였다.

The development of thermoelectric (TE) materials to replace Bi2Te3 alloys is emerging as a hot issue with the potential for wider practical applications. In particular, layered Zintl-phase materials, which can appropriately control carrier and phonon transport behaviors, are being considered as promising candidates. However, limited data have been reported on the thermoelectric properties of metal-Sb materials that can be transformed into layered materials through the insertion of cations. In this study, we synthesized FeSb and MnSb, which are used as base materials for advanced thermoelectric materials. They were confirmed as single-phase materials by analyzing X-ray diffraction patterns. Based on electrical conductivity, the Seebeck coefficient, and thermal conductivity of both materials characterized as a function of temperature, the zT values of MnSb and FeSb were calculated to be 0.00119 and 0.00026, respectively. These properties provide a fundamental data for developing layered Zintl-phase materials with alkali/alkaline earth metal insertions.

PURPOSES : The purpose of this study is to derive dropout rates according to various international roughness index (IRI) specifications using ProVAL, develop a comparative methodology, and indirectly assess the level of road management in each country. METHODS : Based on a literature review, the IRI specifications for each country were collected, and the ProVAL analysis tool was used to compare and analyze dropout rates according to each specification. Thus, the dropout rate rankings for each country were calculated. Additionally, by analyzing the correlation between dropout rates according to each threshold, a model was created to convert the threshold between the most commonly used baseline distances of 100 m and 161 m. RESULTS : Dropout rates were derived according to the standards of each country and rankings were assigned. Comparing 51 standards, the IRI level of New Mexico appeared to be the highest, whereas the domestic specifications ranked 36th. A model was created to convert the threshold between the standard distances of 100 m and 161 m. CONCLUSIONS : This study objectively assessed the roughness standards in various countries using the dropout rate and IRI ranking specifications. The highest specification was found for the asphalt of New Mexico in the USA, with the domestic specification ranking 36th. A model that converts the thresholds between the most commonly used baseline distances of 100 m and 161 m was developed, with slight differences across sections. For a precise conversion, individual models may be required for each section.

Periodontal disease is a chronic but treatable condition which often does not cause pain during the initial stages of the illness. Lack of awareness of symptoms can delay initiation of treatment and worsen health. The aim of this study was to develop and compare different risk prediction models for periodontal disease using machine learning algorithms. We obtained information on risk factors for periodontal disease from the Korea National Health and Nutrition Examination Survey (KNHANES) dataset. Principal component analysis and an auto-encoder were used to extract data on risk factors for periodontal disease. A synthetic minority oversampling technique algorithm was used to solve the problem of data imbalance. We used a combination of logistic regression analysis, support vector machine (SVM) learning, random forest, and AdaBoost to classify and compare risk prediction models for periodontal disease. In cases where we used principal component analysis (PCA) to extract risk factors, the recall was higher than the feature selection method in the logistic regression and support-vector machine learning models. AdaBoost’s recall was 0.98, showing the highest performance of both feature selection and PCA. The F1 score showed relatively high performance in Ada- Boost, logistic regression, and SVM learning models. By using the risk factors extracted from the research results and the predictive model based on machine learning, it will be able to help in the prevention and diagnosis of periodontal disease, and it will be used to study the relationship with various diseases related to periodontal disease.

PURPOSES : This study uses deep learning image classification models and vehicle-mounted cameras to detect types of pavement distress — such as potholes, spalling, punch-outs, and patching damage — which require urgent maintenance. METHODS : For the automatic detection of pavement distress, the optimal mount location on a vehicle for a regular action camera was first determined. Using the orthogonal projection of obliquely captured surface images, morphological operations, and multi-blob image processing, candidate distressed pavement images were extracted from road surface images of a 16,036 km in-lane distance. Next, the distressed pavement images classified by experts were trained and tested for evaluation by three deep learning convolutional neural network (CNN) models: GoogLeNet, AlexNet, and VGGNet. The CNN models were image classification tools used to identify and extract the combined features of the target images via deep layers. Here, a data augmentation technique was applied to produce big distress data for training. Third, the dimensions of the detected distressed pavement patches were computed to estimate the quantity of repair materials needed. RESULTS : It was found that installing cameras 1.8 m above the ground on the exterior rear of the vehicle could provide clear pavement surface images with a resolution of 1 cm per pixel. The sensitivity analysis results of the trained GoogLeNet, AlexNet, and VGGNet models were 93 %, 86 %, and 72 %, respectively, compared to 62.7 % for the dimensional computation. Following readjustment of the image categories in the GoogLeNet model, distress detection sensitivity increased to 94.6 %. CONCLUSIONS : These findings support urgent maintenance by sending the detected distressed pavement images with the dimensions of the distressed patches and GPS coordinates to local maintenance offices in real-time.

최근 다양한 유기재배 작물의 뿌리를 가해하는 굼벵이류의 피해가 증가하고 있으나, 굼벵이류는 토양 내 발생하는 특징으로 인해 발생시기 및 그 종류에 대한 확인이 어려운 해충이다. 피해를 끼치는 굼벵이의 발생을 파악하기 위해 고구마 유기재배지에 페로몬 트랩을 이용하여 굼벵이의 성충의 종류 및 발생 동향을 조사하였다. 조사지는 무안 유기재배농가와 국립농업과학원 완주군 포장에서 이루어졌다. 3종의 풍뎅이 페로몬 루어를 유인제로 사용하였으 며 6월 초부터 8월 말까지 조사지에 트랩을 설치하여 포획된 풍뎅이를 수집하여 동정을 하였다. 유기재배포장에서 포획된 종은 큰검정풍뎅이, 콩풍뎅이, 청동풍뎅이, 녹색콩풍뎅이, 별줄풍뎅이 등의 풍뎅이와 흰점박이꽃무지 등이 주로 채집되었다. 유기재배 고구마포장에서 풍뎅이 발생소장을 조사한 결과 최대로 발생한 시기는 7월초였다.

Japanese pine sawyer beetle, Monochamus alternatus Hope (Coleoptera: Cerambycidae) is considered as a serious pest in pine trees. To develop an eco-friendly strategy to manage this forest insect, we collected entomopathogenic fungi from Korean soil and assessed their virulence against the adults of the insect in laboratory conditions. As a result, two isolates with conidial suspension (1.0×107conidia/ml), showed 87% and 90% mortality 12 days after fungal treatment, respectively. We assessed the potential of the fungi-derived destruxin and protease as additives to the fungal isolates, and they showed insecticidal activity via feeding and spraying treatments. Finally, we produced fungal conidia in massive solid cultures and formulated wettable powders, and now studying optimal conditions of oil-based formulation with two isolates. Based on these results, we are evaluating the control efficacy of the fungal agents against M. alternatus in field conditions.

This study was carried out to investigate insect community structure from different habitats in Baengnyeong island.We performed day and night collection at two different habitats (mountain, rural area) of Baengnyeong island from Mayto September in 2015. A total of 2,879 individuals of 404 species, 81 families belonging to 10 orders were collectedand identified. A dominant species was Idisia ornata Pascoe (Tenebrionidae) despite a very low percentage (6.04%) ofthe species among the catches. Results of independent t-test showed a significant high (p<0.001) of species richness onmountain. Also, seasonal results of ANOVA (Analysis of variance) were significantly influential with species abundanceand species richness. The result of NMDS analysis showed that the community structure of the insects from the mountainis different with rural area.

Recently, N-terminal pro-brain natriuretic peptide (NTproBNP) has been widely used in the areas of diagnosis, monitoring treatment efficiency, and prognosis for various heart diseases, especially heart failure (HF). In this paper, we try to estimate the prognostic significance of NT-proBNP as a risk evaluation marker in Non-ST-segment Elevation Myocardial Infarction (NSTEMI) patients. We selected NSTEMI patients who underwent percutaneous coronary intervention (PCI) primarily using a drug-eluting stent within 24 h after the onset of chest pain. We compared incidences of major adverse cardiac events (MACE) including death, myocardial infarction (MI), stent thrombosis (ST), and target vessel revascularization (TVR) in two patient groups according to a high or low serum concentration of NT-proBNP, which was measured in the emergency room (ER). We intend to minimize selection bias selecting comparing groups, considering covariate of observed variables together using propensity score matching (PSM) and propensity score weighting (PSW) based on propensity score (PS) to control the difference in baseline characteristics between high- and low NT-proBNP groups. We found that as the log NT-proBNP value increases by 1 through a hazard function of COX’s analysis, the risk of MACE increases by 1.312 times. This result indicated that the NT-proBNP level on ER admission can be used as a significant prognostic indicator to estimate 1 year of MACE in NSTEMI patients who were treated with PCI within 24 h after the onset of chest pain.

The spotted-wing drosophila Drosophila suzukii (Diptera: Drosophilidae) is an Asian species introduced into North America and Europe. It damages a wide variety of thin-skinned fruits. We sequenced the complete mitochondrial genome (mitogenome) of D. suzukii to better understand the mitogenomic characteristics of this species and understand phylogentic relationships of Drosophila. The 16,230-bp complete mitogenome of the species consists of a typical set of genes, including 13 protein-coding genes (PCGs), two rRNA genes, and 22 tRNA genes, and one major non-coding A+T-rich region, with an arrangement typical of insects. Twelve PCGs began with the typical ATN codon, whereas the COI began with TCG, which has been designated as the start codon for other Drosophila species. The 1,525-bp A+T-rich region is the second longest in Drosophila species for which the whole mitogenome has been sequenced, after D. melanogaster. Phylogenetic analysis with the 13 PCGs of the Drosophila species using Bayesian Inference and Maximum likelihood methods both placed D. suzukii at the basal lineage of the previously defined Melanogaster group, with a strong support.