딥러닝 기반 영상처리를 이용한 골재 품질 검사
Examination of Aggregate Quality Using Image Processing Based on Deep-Learning
0
조회
0
다운로드
3
인용
0
참고문헌
키워드
초록
콘크리트의 주재료인 골재 중 굵은 골재의 품질관리는 현재 샘플링을 통한 통계적 공정관리(SPC) 방법으로 하고 있다. 본 논문은 굵은 골재에대한 품질관리를 현재의 체거름 방식을 대신 카메라를 통해 획득한 영상을 기반으로 굵은 골재를 검사하게 바꾸어 제조 혁신을 위한 스마트팩토리를 구축하였다. 먼저, 얻은 영상을 전처리 하였고, 딥러닝으로 학습된 HED(Holistically-nested Edge Detection)필터는 각각의 물체를Segmentation하였다. 이 Segmentation한 결과를 영상 처리하여 각각의 골재를 분석 후 이 결과를 바탕으로 조립률, 입형률을 파악한다. 영상을통해 얻은 골재들의 조립률, 입형률을 계산하여 골재의 품질을 검사하였고 알고리즘의 정확도는 실제로 체 가름 방식을 통해 골재의 품질을비교한 것과 90% 이상의 정확도를 보이는 결과가 나왔다. 또한 기존의 방법으로는 골재의 입형률을 검사할 수 없었지만 본문의 내용을 통해골재의 입형률도 측정할 수 있게 되었다. 입형률의 경우 도형을 사용하여 검증하였는데 이는 ±4.5%의 차이를 보였다. 골재의 길이 측정의경우 실제 골재의 길이를 비교하였는데 ±6%의 차이를 보였다. 실제 3차원의 데이터를 2차원의 영상에서 분석하다보니 실제 데이터와 차이가생겼는데 이는 추후 연구가 필요하다.
Abstract
The quality control of coarse aggregate among aggregates, which are the main ingredients of concrete, is currently carried out bySPC(Statistical Process Control) method through sampling. We construct a smart factory for manufacturing innovation by changing thequality control of coarse aggregates to inspect the coarse aggregates based on this image by acquired images through the camera insteadof the current sieve analysis. First, obtained images were preprocessed, and HED(Hollistically-nested Edge Detection) which is the filterlearned by deep learning segment each object. After analyzing each aggregate by image processing the segmentation result, finenessmodulus and the aggregate shape rate are determined by analyzing result. The quality of aggregate obtained through the video was examinedby calculate fineness modulus and aggregate shape rate and the accuracy of the algorithm was more than 90% accurate compared tothat of aggregates through the sieve analysis. Furthermore, the aggregate shape rate could not be examined by conventional methods,but the content of this paper also allowed the measurement of the aggregate shape rate. For the aggregate shape rate, it was verifiedwith the length of models, which showed a difference of ±4.5%. In the case of measuring t```he length of the aggregate, the algorithmresult and actual length of the aggregate showed a ±6% difference. Analyzing the actual three-dimensional data in a two-dimensionalvideo made a difference from the actual data, which requires further research.
참고문헌 (15)
- 류권규, "영상 처리에 의한 하상자갈의 입도분포 분석", 한국수자원학회 논문집, 52(7), pp.493~503 (2019), 한국수자원학회[DOI]
- Y. Kuroda, "Formation mechanism of gravel bed rivers and its characteristics of grain size distribution", Advances in River Engineering, 11, pp.363~368 (2005)[DOI]
- M. Uchio, "Development of a measuring system for particle size distribution in gravel-bed rivers by using an image processing technique", Journal of the Japan Society of Erosion Control Engineering, 58(2), pp.26~31 (2005)[DOI]
- K. Ohashi, "Investigation of riverbed material using image processing", Proceeding of Annual Symposium 2010 of Japan Society of Erosion Control Engineering, pp.286~287 (2010)
- 김용로, "The Quality Status of Aggregate for Domestic Ready-mixed Concrete and the Effect of Aggregate Quality in Concrete", 한국건축시공학회지, 14(1), pp.11~20 (2014), 한국건축시공학회[DOI]
- "Integrated National Standard Certification Information System"[URL]
- B. S. Cho, "An experimental study on the effect of aggregate shape influencing to the concrete properties", Architectural Institute of Korea, Collection of Papers for Academic Presentations, 8(2), pp.735~738 (1988)
- S. Xie, "Holistically-nested edge detection", Proceedings of the IEEE International Conference on Computer Vision, pp.1395~1403 (2015)
- K. Simonyan, "Very deep convolutional networks for large-scale image recognition", ArXiv Preprint ArXiv, ~ (2014)
- N. Silberman, "Indoor segmentation and support inference from rgbd images", (ECCV)European Conference on Computer Vision, pp.746~760 (2012)
- J. Canny, "A computational approach to edge detection", IEEE Transactions on Pattern Analysis and Machine Intelligence, (6), pp.679~698 (1986)
- J. M. Prewitt, "Object enhancement and extraction", pp.15~19 (1970), Picture processing and Psychopictorics
- J. Kittler, "On the accuracy of the sobel edge detector", Image and Vision Computing, 1(1), pp.37~42 (1983)
- "OpenCV Homepage"[URL]
- O. Ronneberger, "U-net : Convolutional networks for biomedical image segmentation", International Conference on Medical Image Computing and Computerassisted Interventio, Springer, Cham, pp, pp.234~241 (2015)