서울 집중호우와 수해

Zheng,Hai-yan,Jin Ri,Kim Dong-joon,Oh,Sung-nam,Lee,Kyoo-seock 한국기상학회 2011 한국기상학회 학술대회 논문집 Vol.2011 No.10-1

Study on Image Augmentation of Leaf Images with Fire Blight Using Paired Dataset and CycleGAN

Ri Zheng,HeLin Yin,Dong Jin,JiMin Lee,Yeong Hyeon Gu,Seong Joon Yoo 한국차세대컴퓨팅학회 2022 한국차세대컴퓨팅학회 학술대회 Vol.2022 No.10

Fire blight is a kind of bacterial disease, which particularly gives serious damage to apples and pears. There is no clear cure for fire blight until now and its infectious speed is fast. Thus, damage due to fire blight should be minimized through early diagnosis. With the development of artificial intelligence in recent years, deep learning has been widely used in the agricultural field. As already known, a deep learning model needs a large number of training datasets. However, fire blight does not occur frequently. Thus, the number of their datasets is very insufficient. To increase this insufficient number of datasets, a data augmentation method in relation to fire blight has been previously conducted but it failed to accurately generate images of fire blight symptoms. In this study, CycleGAN was used to generate accurate fire blight leaf images, and an unpaired dataset, which was used previously by default, was converted into a paired dataset, in which leaves were placed in the same direction. As a result, accurate fire blight leaf images were still not generated when an unpaired dataset was used, but when a paired dataset was used, images with accurate fire blight symptoms were generated.

Characteristics and response of mouse bone marrow derived novel low adherent mesenchymal stem cells acquired by quantification of extracellular matrix

Zheng, Ri-Cheng,Kim, Seong-Kyun,Heo, Seong-Joo,Koak, Jai-Young,Lee, Joo-Hee,Park, Ji-Man The Korean Academy of Prosthodonitics 2014 The Journal of Advanced Prosthodontics Vol.6 No.5

PURPOSE. The aim of present study was to identify characteristic and response of mouse bone marrow (BM) derived low-adherent bone marrow mesenchymal stem cells (BMMSCs) obtained by quantification of extracellular matrix (ECM). MATERIALS AND METHODS. Non-adherent cells acquired by ECM coated dishes were termed low-adherent BMMSCs and these cells were analyzed by in vitro and in vivo methods, including colony forming unit fibroblast (CFU-f), bromodeoxyuridine (BrdU), multi-potential differentiation, flow cytometry and transplantation into nude mouse to measure the bone formation ability of these low-adherent BMMSCs. Titanium (Ti) discs with machined and anodized surfaces were prepared. Adherent and low-adherent BMMSCs were cultured on the Ti discs for testing their proliferation. RESULTS. The amount of CFU-f cells was significantly higher when non-adherent cells were cultured on ECM coated dishes, which was made by 7 days culturing of adherent BMMSCs. Low-adherent BMMSCs had proliferation and differentiation potential as adherent BMMSCs in vitro. The mean amount bone formation of adherent and low-adherent BMMSCs was also investigated in vivo. There was higher cell proliferation appearance in adherent and low-adherent BMMSCs seeded on anodized Ti discs than machined Ti discs by time. CONCLUSION. Low-adherent BMMSCs acquired by ECM from non-adherent cell populations maintained potential characteristic similar to those of the adherent BMMSCs and therefore could be used effectively as adherent BMMSCs in clinic.

AI 기반 문화 빅데이터 분석 스튜디오 설계 및 구현

Ri Zheng,HeLin Yin,Jae Yoo Lee,Yeong Hyeon Gu,Seong Joon Yoo 한국차세대컴퓨팅학회 2021 한국차세대컴퓨팅학회 학술대회 Vol.2021 No.05

최근 사람들은 Artificial Intelligent (AI) 기반 빅데이터 분석 시스템에 많은 관심을 가지고 있고 자신만의 빅데이터 모델을 원한다. 하지만 이들은 데이터가 있음에도 불구하고 프로그래밍 지식, 컴퓨팅 파워, 인공지능 경험 등 제한으로 인해 해당 빅데이터 데이터를 분석하기 어려운 문제점이 있다. 본 연구에서는 이러한 문제점을 해결하기 위해 웹(web) 기반 문화 빅데이터 분석 모델 개발 스튜디오를 제안한다. 제안한 시스템은 정형 데이터 분석 모듈, 토픽 모델링 모듈, 전통문양 검색 모듈 3개 부분으로 구성되었다. 정형 데이터 분석 모듈에서는 사용자가 가지고 있는 문화 정형 데이터를 기계학습을 사용해 분석할 수 있다. 토픽 모델링 모듈은 사용자가 가지고 있는 문화 텍스트 데이터를 기계학습 및 자연어처리 기술을 사용해 해당 데이터의 문서 집합의 추상적인 주제를 추출할 수 있다. 전통문양 검색 모듈에서는 사용자가 가지고 있는 전통문양 이미지를 입력해 분석할 이미지의 10개 유사 이미지를 추출할 수 있다. 사용자는 필요한 모듈을 선택해 자신만의 데이터로 문화 빅데이터 모델을 구현할 수 있을 뿐만 아니라 모델의 성능도 측정할 수 있다.

Helicobacter Pylori CagA and Gastric Carcinogenesis

Zheng, Ri-Nan,Li, Shu-Rong,Masahiro, Asaka Asian Pacific Journal of Cancer Prevention 2012 Asian Pacific journal of cancer prevention Vol.13 No.12

Objectives: This study aimed to demonstrate the tyrosine phosphorylation motif (TPM) and 3' region structure of the Helicobacter pylori CagA gene as well as its SHP-2 binding activity in AGS cells and relation to gastric carcinogenesis. Methods: Sixteen clinical isolate H. pylori strains from eight duodenal ulcer and eight gastric adenocarcinoma patients were studied for CagA repeat sequence EPIYA motifs, C-terminal structure, and western blot analysis of CagA protein expression, translocation, and SHP-2 binding in AGS cells. Results: Except for strain 547, all strains from the gastric adenocarcinoma patients were positive for CagA by PCR and had three EPIYA copy motifs. Western blotting showed that all strains were positive for CagA protein expression (100%), CagA protein translocation (100%), and SHP-2 binding (100%). CagA protein expression was significantly higher in the gastric adenocarcinoma patients than in the duodenal ulcer patients (P=0.0023). CagA protein translocation and SHP-2 binding in the gastric adenocarcinoma patients were higher than those in the duodenal ulcer patients, but no significant differences were found between the two groups (P=0.59, P=0.21, respectively). Conclusions: The TPMs and 3' region structures of the H. pylori CagA gene in the duodenal ulcer and gastric adenocarcinoma patients have no significant differences.

유사도 기반 병해충 검색 모델에 대한 검증 연구

Ri Zheng,Dong Jin,Helin Yin,구영현,유성준 한국차세대컴퓨팅학회 2022 한국차세대컴퓨팅학회 학술대회 Vol.2022 No.05

농작물 병해충은 심각한 농업재해 중 하나로 농작물의 생산량과 품질에 큰 영향을 미치고 있다. 이는 병해충의 예방과 방제를 통해 농장에 대한 손해와 경제적 손실을 최소화할 수 있다. 병해충의 예방과 방제를 위해 최근 딥러닝 기반 병해충 인식 연구들이 많이 진행되고 있다. 그 중 대부분 연구에서는 분류(Classification) 기법을 주로 사용하고 있다. 분류 기법에서는 가장 확률이 높은 값을 가지는 하나의 클래스를 출력한다. 하지만 분류 모델의 성능이 100%가 아니기 때문에 충분히 잘못된 결과를 출력할 수 있다. 이러한 문제점을 해결하기 위해 Yin et al. [1]에서 여러 장의 병해충 이미지를 출력하는 유사도 기반 병해충 이미지 검색 모델을 제안했다. 본 연구에서는 Yin et al. [1]에서 제안한 병해충 이미지 검색 모델을 사과, 배추, 감귤 등 10종 작물, 50종 병해충 데이터 셋에 적용해 모델의 성능을 검증했다. 실험 측정 결과 병해충 이미지 검색 모델은 약 83.20%~99.71%의 유사 정확도를 보였다.

과수화상병 이미지 수집 관리 시스템 구현

Ri Zheng,이지민(Ji Min Lee),Dong Jin,Helin Yin,유성준(Seong Joon Yoo),구영현(Yeong Hyeon Gu) 대한전자공학회 2023 대한전자공학회 학술대회 Vol.2023 No.6

Fire blight is a bacterial disease that is particularly damaging to apples and pears. There is no clear treatment method for fire blight until now, and its infectious speed is fast. Thus, damage due to fire blight should be minimized through early diagnosis. Recently, with the development of artificial intelligence technology, deep learning technology is widely applied in the agricultural field. As already known, a deep learning model needs a large number of training datasets. However, fire blight does not occur frequently. Thus, the number of their datasets is very insufficient. To increase this insufficient number of datasets, we propose a fire blight image collection and management system to build a high-quality fire blight training dataset. The proposed system is largely composed of three modules: an APP-based fire blight image collection module, an image inspection module, and a data annotation module. First, images of fire blight were collected across the country. Then inspect the collected images to filter out images that are not fire blight. Finally, the fire blight images are annotated to generate a high-quality artificial intelligence learning dataset necessary for fire blight recognition.

농작물 이미지 판별 모델에 대한 연구

Ri Zheng,Helin Yin,구영현(Yeong Hyeon Gu),유성준(Seong Joon Yoo) 대한전자공학회 2021 대한전자공학회 학술대회 Vol.2021 No.11

도시廣場의 장소성에서 나타난 도시정체성에 관한 연구

정은일(Zheng En-Ri),양영준(Yang Young-Joon) 대한건축학회 2011 대한건축학회논문집 Vol.27 No.4

Square is the most important public space in the city. It always has special relationship with the city. And it relationship with place. Place with several meanings represent the identity of the city. The research from this perspective, place and identity in the plaza of the city is to clarify the relationship in Xi'an that the most famous historic city in China. To clarify this relationship, that place would be divided historical significance and cultural meaning, And identity of city would be divided historical context and the local cultural context.

廣場의 도시적 기능과 의미에 관한 연구

정은일(Zheng En-Ri),양영준(Yang Young-Joon),이상준(Lee Sang-Jun) 대한건축학회 2010 대한건축학회논문집 Vol.26 No.10

Thousands of years, the plaza is a genre of human civilization history. The plaza that the most important part of the city, was changing and developing constant dynamically with city transition. The passed half of century, China constructed a lot of emerging cities. These emerging cities constructed a lot of plazas for finding the solution of shortage public spaces. Changchun that one of the most important industrial city in China, constructions many plazas. Through this research, finding urban function and meaning of plaza will be useful in planning plaza to urban construction in future.