엣지 컴퓨팅 환경에서 적용 가능한 딥러닝 기반 라벨 검사 시스템 구현
Implementation of Deep Learning-based Label Inspection System Applicable to Edge Computing Environments
0
조회
0
다운로드
2
인용
0
참고문헌
키워드
초록
In this paper, the two-stage object detection approach is proposed to implement a deep learning-based label inspection system on edge computing environments. Since the label printed on the products during the production process contains important information related to the product, it is significantly to check the label information is correct. The proposed system uses the lightweight deep learning model that able to employ in the low-performance edge computing devices, and the two-stage object detection approach is applied to compensate for the low accuracy relatively. The proposed Two-Stage object detection approach consists of two object detection networks, Label Area Detection Network and Character Detection Network. Label Area Detection Network finds the label area in the product image, and Character Detection Network detects the words in the label area. Using this approach, we can detect characters precise even with a lightweight deep learning models. The SF-YOLO model applied in the proposed system is the YOLO-based lightweight object detection network designed for edge computing devices. This model showed up to 2 times faster processing time and a considerable improvement in accuracy, compared to other YOLO-based lightweight models such as YOLOv3-tiny and YOLOv4-tiny. Also since the amount of computation is low, it can be easily applied in edge computing environments.
Abstract
In this paper, the two-stage object detection approach is proposed to implement a deep learning-based label inspection system on edge computing environments. Since the label printed on the products during the production process contains important information related to the product, it is significantly to check the label information is correct. The proposed system uses the lightweight deep learning model that able to employ in the low-performance edge computing devices, and the two-stage object detection approach is applied to compensate for the low accuracy relatively. The proposed Two-Stage object detection approach consists of two object detection networks, Label Area Detection Network and Character Detection Network. Label Area Detection Network finds the label area in the product image, and Character Detection Network detects the words in the label area. Using this approach, we can detect characters precise even with a lightweight deep learning models. The SF-YOLO model applied in the proposed system is the YOLO-based lightweight object detection network designed for edge computing devices. This model showed up to 2 times faster processing time and a considerable improvement in accuracy, compared to other YOLO-based lightweight models such as YOLOv3-tiny and YOLOv4-tiny. Also since the amount of computation is low, it can be easily applied in edge computing environments.
참고문헌 (25)
- 강병두, "주성분 분석과 서포트 벡터 머신을 이용한 효과적인 얼굴 검출 시스템", 멀티미디어학회논문지, 9(11), pp.1435~1444 (2006), 한국멀티미디어학회
- 맹형열, "다중 스펙트럼 머신비전 응용을 위한 CUDA SURF 기반의 영상 정렬 기법", 멀티미디어학회논문지, 17(9), pp.1041~1051 (2014), 한국멀티미디어학회[DOI]
- 홍승범, "머신비전 자동검사를 위한 대상객체의 인식방향성 개선", 한국정보통신학회논문지, 23(11), pp.1384~1390 (2019), 한국정보통신학회[DOI]
- 김호연, "사출성형 자동차부품의 외관 비전검사 방법", 한국정보통신학회논문지, 23(2), pp.127~132 (2019), 한국정보통신학회[DOI]
- 윤석문, "다각형 용기의 결함 검사 시스템 개발", 전기전자학회논문지, 25(3), pp.485~492 (2021), 한국전기전자학회
- 김민재, "전이학습을 이용한 비전기반 결함 탐지 알고리즘 개발", 한국CDE학회 논문집, 25(3), pp.246~255 (2020), 한국CDE학회[DOI]
- 강건하, "오토인코더를 사용한 이상탐지 모델의 비교분석 및 이상치 판별 기준 제안", 한국컴퓨터정보학회논문지, 26(8), pp.23~30 (2021), 한국컴퓨터정보학회[DOI]
- 박혜진, "심층학습 기법을 활용한 효과적인 타이어 마모도 분류 및 손상 부위 검출 알고리즘", 멀티미디어학회논문지, 24(8), pp.1026~1034 (2021), 한국멀티미디어학회
- S. Ren, "Faster R-Cnn:Towards Real-Time Object Detection with Region Proposal Networks", IEEE Transactions on Pattern Analysis and Machine Intelligence, 39(6), pp.1137~1149 (2016)
- J. Redmon, "You Only Look Once: Unified, Real-Time Object Detection", Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pp.779~788 (2016)
- P. Adarsh, "YOLO v3-Tiny:Object Detection and Recognition using One Stage Improved Model", 2020 6th International Conference on Advanced Computing and Communication Systems, pp.687~694 (2020)
- B. G. Han, "Design of a Scalable and Fast YOLO for Edge-Computing Devices", Sensors, 20(23) (2020)
- 정유수, "소형 머신 비전 검사 장비에 기반한 O링치수 측정", 한국산업정보학회논문지, 19(4), pp.41~52 (2014), 한국산업정보학회[DOI]
- 박병준, "색상 검사와 템플릿 매칭을 이용한 정션 박스 자동 시각 검사 시스템", 멀티미디어학회논문지, 13(3), pp.392~399 (2010), 한국멀티미디어학회
- 고동민, "고정밀 머신 비전을 위한 정확한 PCB 윤곽선과 코너 검출", 반도체디스플레이기술학회지, 16(3), pp.53~58 (2017), 한국반도체디스플레이기술학회
- 김기석, "SVM 학습 알고리즘을 이용한 자동차 썬루프의 부품 유무 비전검사 시스템", 제어.로봇.시스템학회 논문지, 19(12), pp.1099~1104 (2013), 제어·로봇·시스템학회[DOI]
- 김대현, "딥러닝 알고리즘을 이용한 머신 비전 기반 불량 검출 연구", 비파괴검사학회지, 40(1), pp.47~52 (2020), 한국비파괴검사학회[DOI]
- 류가애, "영상 데이터 기반의 CNN을 이용한 제조 공정 데이터 분류 적용에 대한 연구", 정보화연구, 15(3), pp.337~343 (2018), 한국엔터프라이즈아키텍처학회[DOI]
- R. Girshick, "Fast r-cnn", Proceedings of the IEEE International Conference on Computer Vision, pp.1440~1448 (2015)
- K. He, "Mask R-Cnn", Proceedings of the IEEE International Conference on Computer Vision, pp.2961~2969 (2017)
- J. Redmon, "Yolov3: An Incremental Improvement", Computer Vision and Pattern Recognition, pp.1804~2767 (2018)
- A. Bochkovskiy, "Yolov4:Optimal Speed and Accuracy of Object Detection", ~ (2020)
- C. Y. Wang, "Scaled-Yolov4 : Scaling Cross Stage Partial Network", Proceedings of the IEEE, 2021, 13029~13038
- Y. Li, "YOLOv3-lite: A Lightweight Crack Detection Network for Aircraft Structure Based on Depthwise Separable Convolutions", Applied Sciences, 9(18) (2019)