http://chineseinput.net/에서 pinyin(병음)방식으로 중국어를 변환할 수 있습니다.
변환된 중국어를 복사하여 사용하시면 됩니다.
- 中文 을 입력하시려면 zhongwen을 입력하시고 space를누르시면됩니다.
- 北京 을 입력하시려면 beijing을 입력하시고 space를 누르시면 됩니다.
RISS 인기검색어
- 검색키워드
- 저자 : 이현수 (검색결과 268 건)
-
-
초점거리에 따른 배광분석 및 색수차 분석을 통한 자동차용 LED 헤드램프의 성능향상에 관한 연구
LED as a light source is making big changes in automotive lighting field. The optical system has been changed to adopt a LED. For example in the engineering field, we have used a heatsink and FAN to dissipate a heat in the LED optical system. Before we have used a LED, we had just adopted a natural convection instead of forced convections. And also style of headlamp has a big change. Normally headlamp has a big circle image to express an eye of vehicle. But now, LED headlamps express their own design characteristic by using different looks. LED gives us many chances to use putting plastic materials into the headlamps. One of big chances is a temperature in the system. Before using LED, traditional light sources has high temperature. That is why we have to use a glass to make an optical lens into a optical system. LED optical system could use a plastic lens. it means we can make almost all kinds of shapes through the plastic injection molding. Already some car manufactures are adopting this technology into their design to express to all customer their own identity. Especially premium automotive companies. To use this technology, we have to overcome a big defect to adopt this technology. That is a 'Color abberation' near the cut off line of low beam. First of all, I am looking forward to understand why we need this technology in automotive field. so I will describe main factors about the automotive optical system in this thesis. After that, I have to study what is the problem and find causes. Then, I will analyze and find a new solution.
-
한글 폰트를 변형하기 위한 비지도 학습 모델 기반 폰트 속성 제어 연구
한글은 로마자에 비해 11,172자라는 많은 글자 때문에 폰트 제작에 많은 시간과 비용이 든다. 이를 해결하기 위해 기존에 외곽선 폰트와 METAFONT를 이용하여 폰트를 변형하는 연구가 진행되었으나 데이터 처리에 시간이 많이 드는 문제가 있었다. 따라서 인공지능을 이용 자동으로 폰트를 생성하는 연구가 제안되었다. 이는 지도 학습 모델을 이용하여 적은 글자로 한글 폰트 한 세트를 만드는 방향으로 진행되었다. 그러나 지도학습 모델은 비지도 학습 모델에 비해 다양한 스타일를 가진 폰트를 생성 할 수 없다. 따라서 본 논문에서는 비지도 학습 모델인 StyleGAN 모델 구조에 인코더를 추가한 FMGAN 구조를 제안하고 FMGAN의 Font Mixing 방법과 모델의 잠재공간을 이용하여 한글 폰트의 여러 속성을 제어하는 걸 보인다. 제안하는 방법을 통해 비지도 학습 모델과 잠재 공간을 제어 할 수 있는 가능성을 보인다. Because Hangul has 11,172 characters compared to Roman characters, it takes a lot of time and money to make fonts. In order to solve this problem, research has been conducted to transform the font using the outline font and METAFONT, but there is a problem that data processing takes a lot of time. Therefore, research on automatically generating fonts using artificial intelligence has been proposed. This proceeded in the direction of making a set of Hangul fonts with small letters using a supervised learning model. However, the supervised learning model cannot generate fonts with different styles compared to the unsupervised learning model. Therefore, in this paper, we propose an FMGAN structure with an encoder added to the StyleGAN model structure, which is an unsupervised learning model, and show how to control various properties of Korean fonts using FMGAN's Font Mixing method and the latent space of the model. The proposed method shows the possibility of controlling unsupervised learning models and latent spaces.
-
정신지체 학생과의 통합체육수업에 대한 비장애학생의 태도가 관여도 및 상호작용행동에 미치는 영향
The purpose of this study was to find out the difference in the interaction behavior, involvement and attitude in inclusive physical education of non-disabilities who are very important for adaptation of disabled students to inclusive physical education. Also it was to find out the relationship among the interaction behavior, involvement and the attitude of non-disabilities in inclusive physical education. The population of this study was non-disabilities who experienced the inclusive physical education with disabled students in Busan. Through purposeful sampling, 700 students without disabilities were asked for responding the survey. Among 700 data, 668 data were finally used for this study. The questionnaire was put together with 5 questions of demographic factor, 20 questions of attitude in inclusive physical education, 5 questions of involvement in inclusive physical education and 9 questions of interaction behavior. The questions of non-disabilities attitude in inclusive physical education were revised and based on Oh(2001b) and the questions of involvement in inclusive physical education were revised and based on Lee(2004b, 2005c). And the questions of interaction behavior were revised and based on Kim(2000), Lee(2004b, 2005c), Lee and Nam(2004) and Ha(2000). To analyze the data collected for this study, SPSS 11.0 was used for exploratory factor analysis, t-test and One-way ANOVA analysis. Also, Amos 4.01 was used for confirmatory factor analysis, multiple group factor analysis and structural equation model analysis. Conclusion of this study are as follows. 1) Difference in the attitude, involvement, interaction behavior First, there were no difference in the attitude, involvement for inclusive physical education and the interaction behavior of non-disabilities according to gender. And the result showed that girls' interaction behavior with general disabled people was higher than boys'. Second, there were no differences in the attitude, involvement for inclusive physical education and the interaction behavior of non-disabilities according to the experience period of inclusive physical education. And the result shows that as the period was longer, interaction behavior with general disabled people was higher. Third, there were no differences in the attitude for inclusive physical education even they have any disabled relatives or family members and the result shows that the involvement in inclusive physical education, interaction behavior with disabled students and interaction behavior with general disabled people of non-disabilities who have a disabled relative of family member was higher. Fourth, the attitude, involvement for inclusive physical education and interaction behavior with general disabled people of non-disabilities who had disabled friends were higher than non-disabilities who didn't have any disabled friends. Fifth, the attitude, involvement for inclusive physical education and interaction behavior with general disabled people of non-disabilities who had ever experienced being a partner with disabled students were higher than non-disabilities who had never experienced being a partner with disabled students. Sixth, the attitude, involvement for inclusive physical education and interaction behavior with general disabled people of non-disabilities who participate in physical education class with disabled students were higher than non-disabilities who distract the class and outcasts, or who behave individually in class. 2) The related effects on the attitude, involvement, interaction behavior First, the attitude of non-disabilities for mental retarded students in Inclusive Physical Education showed positive effects on involvement in inclusive physical education. Second, the attitude of non-disabilities for mental retarded students in inclusive physical education showed positive effects on interaction behavior with mental retarded students in inclusive physical education. Third, the involvement of inclusive physical education showed positive effects on interaction behavior with mental retarded students in inclusive physical education. Fourth, the involvement of inclusive physical education showed positive effects on interaction behavior with general disabled people. Fifth, the interaction behavior with mental retarded students in inclusive physical education showed positive effects on interaction behavior with general disabled people.
-
-
영역별 오차 특성을 고려한 IMM-KF 기반 레이더 센서 융합
최근 지능형 자동차 연구에는 다양한 센서를 이용한 첨단 운전자 지원 시스템 연구가 활발히 진행 되고 있으며, 신뢰도 높은 센서는 자율주행 기술에 있어 필수 불가결한 요소이다. 레이더는 타 센서들에 비해 외부환경에 강인한 특성으로 인해 다양한 능동 안전 시스템 연구에 적용되고 있고, 대표적 양산 기술로는 후방 사각지대 경보 시스템을 들 수 있다. 그러나 상용 레이더 역시 작동 메커니즘과 트래킹 방식에 따른 영역별 오차가 존재하며, 후방 사각지대 감지 용도로 레이더를 사용할 경우, 옆 차선과 옆옆차선의 A,B,C 필라 부근에서의 감지 공백 영역이 생긴다. 본 논문에서는 상용 레이더(24GHz)를 이용한 후방 사각 영역 감지 시스템에서 발생하는 감지 공백 영역과, 상용 레이더의 오차를 보완하기 위해 영역별 오차 특성을 고려한 Interacting Multiple Model Kalman Filter(IMM-KF) 기반 센서 융합 기법을 제안한다. 먼저 상용 레이더의 오차를 줄이기 위해, 상대 차량과의 상대거리를 기준으로 균일하게 나눈 영역별 오차 특성을 실험적으로 구한 뒤, IMM-KF를 전 측방 레이더, 후 측방 레이더 각각에 적용함으로써 트래킹의 정확도를 높였다. 이 후 자 차량 주변 사각영역에 대한 감지 영역 향상을 위해 후 측방 레이더와 전 측방 레이더의 센서 융합을 통해 Ground Truth로 사용하는 Differential Global Positioning System (DGPS)에 가까운 융합 결과를 얻을 수 있었다. 제안된 알고리즘은 실험 차량을 통해 취득한 Radar 값과 DGPS 값을 토대로 분석 및 검증되었다.
-
-
-
-
연관 검색어 추천
이 검색어로 많이 본 자료
활용도 높은 자료
