• Smart Structures and Systems
• /
• v.29 no.1
• /
• pp.195-206
• /
• 2022

Crack detection is essential for inspection of existing structures and crack segmentation based on deep learning is a significant solution. However, datasets are usually one of the key issues. When building a new dataset for deep learning, laborious and time-consuming annotation of a large number of crack images is an obstacle. The aim of this study is to develop an approach that can automatically select a small portion of the most informative crack images from a large pool in order to annotate them, not to label all crack images. An active learning method with difficulty learning mechanism for crack segmentation tasks is proposed. Experiments are carried out on a crack image dataset of a steel box girder, which contains 500 images of 320×320 size for training, 100 for validation, and 190 for testing. In active learning experiments, the 500 images for training are acted as unlabeled image. The acquisition function in our method is compared with traditional acquisition functions, i.e., Query-By-Committee (QBC), Entropy, and Core-set. Further, comparisons are made on four common segmentation networks: U-Net, DeepLabV3, Feature Pyramid Network (FPN), and PSPNet. The results show that when training occurs with 200 (40%) of the most informative crack images that are selected by our method, the four segmentation networks can achieve 92%-95% of the obtained performance when training takes place with 500 (100%) crack images. The acquisition function in our method shows more accurate measurements of informativeness for unlabeled crack images compared to the four traditional acquisition functions at most active learning stages. Our method can select the most informative images for annotation from many unlabeled crack images automatically and accurately. Additionally, the dataset built after selecting 40% of all crack images can support crack segmentation networks that perform more than 92% when all the images are used.

• Archives of Plastic Surgery
• /
• v.47 no.5
• /
• pp.428-434
• /
• 2020

Background Three-dimensional (3D) model printing improves visualization of anatomical structures in space compared to two-dimensional (2D) data and creates an exact model of the surgical site that can be used for reference during surgery. There is limited evidence on the effects of using 3D models in microsurgical reconstruction on improving clinical outcomes. Methods A retrospective review of patients undergoing reconstructive breast microsurgery procedures from 2017 to 2019 who received computed tomography angiography (CTA) scans only or with 3D models for preoperative surgical planning were performed. Preoperative decision-making to undergo a deep inferior epigastric perforator (DIEP) versus muscle-sparing transverse rectus abdominis myocutaneous (MS-TRAM) flap, as well as whether the decision changed during flap harvest and postoperative complications were tracked based on the preoperative imaging used. In addition, we describe three example cases showing direct application of 3D mold as an accurate model to guide intraoperative dissection in complex microsurgical reconstruction. Results Fifty-eight abdominal-based breast free-flaps performed using conventional CTA were compared with a matched cohort of 58 breast free-flaps performed with 3D model print. There was no flap loss in either group. There was a significant reduction in flap harvest time with use of 3D model (CTA vs. 3D, 117.7±14.2 minutes vs. 109.8±11.6 minutes; P=0.001). In addition, there was no change in preoperative decision on type of flap harvested in all cases in 3D print group (0%), compared with 24.1% change in conventional CTA group. Conclusions Use of 3D print model improves accuracy of preoperative planning and reduces flap harvest time with similar postoperative complications in complex microsurgical reconstruction.

• Toxicological Research
• /
• v.23 no.4
• /
• pp.353-362
• /
• 2007

Deoxynivalenol (DON) is a common food borne mycotoxin and occurs predominantly in grains such as wheat, barley, oats, etc. DON induces systemic health problems such as loss of appetite, emesis and diarrhea in both human and farm animals. Reliable diagnostic parameters for DON intoxication are needed to prevent deep health impact. In order to establish useful diagnostic parameters, we investigated clinical signs, hematological values, serum biochemical values, gross-, histo- and toxico-pathological findings in B6C3F1 male mice after oral administration of DON (0.83, 2.5 and 7.5 mg/kg) for 8 days. Body weight gain was significantly decreased at the highest dose of DON. Anorexia, ataxia, for crudness and lack of vigor were observed at the highest dose DON group. In hematological values, the numbers of WBC and platelets and hemoglobin content were reduced with decreased neutrophil and monocytes by 7.5 mg/kg DON. Prothrombin time (PT) and activated partial thromboplastin time (aPTT) were prolonged in a dose-dependent manner and the content of fibrinogen was elevated at high dose of DON. Of serum biochemical values, total protein, globulin, BUN, cholesterol and test-osterone were reduced but total bilirubin and albumin/globulin ratio increased. The enzyme activity of alkaline phosphatase was decreased while that of alanine aminotransferase was elevated. Relative organ weights of thymus, seminal vesicle/prostate and testes were dose-dependently reduced but those of liver and left adrenal gland increased with dose dependency. As for pathological findings, atrophy of thymus, seminal vesicle/prostate and testes and submucosal edema and ulceration in stomach and depletion of lymphocytes in thymus cortex were observed. In conclusion, these clinical, hematological, blood biochemical and patholgical parameters obtained in the present studies can be used for diagnosis of DON-mycotoxicosis, especially, low WBC, platelets, protein, BUN and testosterone and delayed prothrombin time can be available as for reliable diagnostic parameters.

• Journal of the Institute of Electronics Engineers of Korea SD
• /
• v.41 no.12
• /
• pp.27-34
• /
• 2004

The study on the RF substrate circuit is necessary to model RF output characteristics of deep submicron MOSFETs below 0.2$\mum$ gate length that have bun commercialized by the recent development of Si submicron process. In this paper, direct extraction methods are developed to apply for a simple substrate resistance model as well as another substrate model with connecting resistance and capacitance in parallel. Using these extraction methods, better agreement with measured Y22-parameter up to 30 GHz is achieved for 0.15$\mum$ CMOS device by using the parallel RC substrate model rather than the simple resistance one, demonstrating the RF accuracy of the parallel model and extraction technique. Using this model, bias and gate length dependent curves of substrate parameters in the RF region are obtained by increasing drain voltage of 0 to 1.2V at deep submicron devices with various gate lengths of 0.11 to 0.5㎛ These new extraction data will greatly contribute to developing a scalable RF nonlinear substrate model.

• The Bulletin of The Korean Astronomical Society
• /
• v.40 no.2
• /
• pp.41.2-41.2
• /
• 2015

We have performed deep wide-field CCD photometry of the young open cluster IC 1805 in the famous star forming region W4, and obtained photometric data for more than 91,000 stars in the field of IC 1805 based on observations with the 3.6m CFHT and the AZT-22 1.5m telescope at Maidanak Astronomical Observatory in Uzbekistan. The photometric data cover an area $43^{\prime}{\times}45^{\prime}$ which is far larger and far deeper than any other optical observations made for the cluster. In order to select the young stellar objects with mid-IR excess emission, we have performed mid-IR photometry of the cluster using the archival images obtained with the Spitzer Space Telescope IRAC and MIPS instruments. From a preliminary analysis of the data, we determined the reddening law ($R_V=3.02{\pm}0.05$), distance modulus ($V_0-M_V=11.9{\pm}0.2$), and the spatial distribution of members.

• 한국정보디스플레이학회:학술대회논문집
• /
• 2003.07a
• /
• pp.1147-1150
• /
• 2003

When a braun tube becomes wider, one of the major problems to be experienced is the non-uniform coating along the four diagonal directions on its surface. This non-uniformity in the coating thickness has a deep relation with the fluid flow on the surface of a braun tube. In order to control the fluid flow properly, we install the plate to block fluid flow at the corner of a braun tube. In the present study, we investigate the effects of the geometry of plate to control the fluid flow and coating uniformity and determine the optimal shape of plate to improve the quality of coating uniformity.

• Proceedings of the Computational Structural Engineering Institute Conference
• /
• 1991.04a
• /
• pp.8-14
• /
• 1991

The finite element technique incorporating infinite elements is applied to analyzing the general three dimensional wave-structure interaction problems within the limits of linear wave theory. The hydrodynamic farces are assumed to be inertially dominated, and viscous effects are neglected. In order to analyze the corresponding boundary value problems efficiently, two types of elements are developed. One is the infinite element for modeling the radiation condition at infinity, and the other is the fictitious bottom boundary element for the case of deep water. To validate those elements, numerical analyses are performed for several floating structures. Comparisons with the results from culler available solution methods show that the present method incorporating tile infinite and the fictitious bottom boundary elements gives good results.

• Proceedings of the Korean Society of Tribologists and Lubrication Engineers Conference
• /
• 1998.10a
• /
• pp.181-189
• /
• 1998

In order to investigate the effect of base oil composition on the electronic insulating oil's performances, an experimental study has been conducted using different oils. Owing to their properties, like lower pour point and gas absorbing, naphthenic base oils are used more often than paraffmic base oils for the electronic insulating oil application. Naphthenic and paraffinic base oils are significantly different in their aromatic hydrocarbon content. In this paper, PXE(para xylyl ethane), LAB(linear alkylbenzene), C13 aromatic hydrocarbon mixture and C17 aromatic hydrocarbon mixture are investigated regarding their influence on insulating oil's performances. According to present study, breakdown voltage decreased with increasing aromatic lydrocarbon content in a deep dewaxed paraffmic base oil. However, any changing in the dissipation factor was not recognizable at small treated level. Furthermore, the volume resistance was not influenced by aromatic hydrocarbon content. The gassing tendency was found as a highly sensible property, changing with treating aromatic hydrocarbons. The higher benzene ring content in the hydrocarbon, the better gassing tendency.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2022.05a
• /
• pp.438-440
• /
• 2022

산업 현장에서 인력으로 작업할 수 있는 물리적 한계를 극복하기 위해 특수 목적 차량 작업 시 차량의 넘어짐 방지와 차체 보호를 위해 아웃트리거를 착지시키는데 도로 상태에 따라 사용자가 직접 최적화를 수행하는데 어려움이 존재한다. 본 논문에서는 도로 표면 상태를 신속하게 판단하여 아웃트리거 수직 및 수평 전개 착지 시 시간 소모, 안전사고 발생을 낮추기 위해 시멘틱 분할을 이용한 도로 표면 상태를 분석하는 연구를 수행하였다. 13가지로 구분된 도로 표면 상황에 대하여 DeepLabV3+를 통해 실험한 결과 픽셀 성능0.7819, mIoU 0.7085 결과를 도출하였다.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2022.11a
• /
• pp.592-593
• /
• 2022

다중 데이터 기반의 행동 인식 과정에서 데이터 수집 반경이 비교적 제한되는 영상 데이터의 결측에 대한 보완이 요구된다. 본 논문에서는 6축 센서 데이터를 이용하여 결측된 영상 데이터를 생성함으로써 행동 인식의 성능을 개선하는 방법을 제안한다. 가속도와 자이로 센서로부터 수집된 행동 데이터를 이용하여 GAN(Generative Adversarial Network)을 통해 영상에서의 관절(Skeleton) 움직임에 대한 데이터를 생성하고자 한다. 이를 위해 DeepLabCut 기반 모델 학습을 통해 관절 좌표를 추출하며, 전처리된 센서 시퀀스 데이터를 가지고 GRU 기반 GAN 모델을 통해 관절 좌표에 대한 영상 시퀀스 데이터를 생성한다. 생성된 영상 시퀀스 데이터는 영상 데이터의 결측이 발생했을 때 대신 행동 인식 모델의 입력값으로 활용될 수 있어 성능 향상을 기대할 수 있다.