| 논문명 | 실내이미지 자동인식 기술을 활용한 지능형 디자인 사례 관리 - 국내 아파트 디자인 사례를 중심으로 -/A Approach to an Intelligent Management of Interior Design References Pictures using Auto-recognition Technique - Focused on the Rooms of Korean Apartments -/3-학술발표1 |
|---|---|
| 저자명 | 김진성(Kim, Jin-Sung) ; 이진국(Lee, Jin-Kook) |
| 발행사 | 한국실내디자인학회 |
| 수록사항 | 한국실내디자인학회 학술발표대회 논문집 , 제20권 2호 |
| 페이지 | 시작페이지(72) 총페이지(4) |
| 주제분류 | 계획및설계 |
| 주제어 | 실내디자인 사례 이미지 ; 딥러닝 ; 이미지 자동인식 ; 지능형 사례 관리 Interior Design Reference Images ; Deep Learning ; Image Automatic Recognition ; Intelligent Reference Management |
| 요약2 | This paper aims to propose an approach to intelligent management system of interior design reference pictures using deep learning-based image recognition technique and demonstration of automated recognition and classification of the picture. that play an important role in the approach. In general interior design process, the pictures of precedent design cases could be very useful as a reference for both designers and clients. it is important to efficiently collect, classify, save and search such reference data, but the current approaches rely on manual works so it takes lots of time and human resource. Recently, it is possible to train the computer to recognize and classify digital images regarding specific criteria using deep learning algorithm. Using this technique and pre-trained computer model, we retrain the model to recognize the Korean apartment room design and classify room usage(e.g. bathroom, bedroom, entrance, kitchen and living room) of the images. the results of test the model is summarized with correct answer rate and accuracy. Using the model, the automated classification application is demonstrated with GUI interface. |
| 소장처 | 한국실내디자인학회 |