택배 배송 서비스 품질향상을 위한 YOLO 기반 택배 상자 파손 탐지 및 DCGAN 기반 데이터 증강
YOLO-based Detection of Damaged Parcel Boxes and Data Augmentation Using DCGAN for Enhancement of Parcel Delivery Service Quality
0
조회
0
다운로드
0
인용
0
참고문헌
키워드
초록
The damage to parcel not only diminishes customer satisfaction with parcel delivery services but also incurs return logistics costs, and this problem can be addressed by detecting damage before parcels reach the customers. Therefore, various deep learning researches related to the parcel damage detection are being conducted. In this study, we introduce a method for parcel box damage detection that not only identifies the presence of damage but also determines its specific type. To achieve this, we employ DCGAN-based data augmentation and use the YOLOv5s-cls and YOLOv5m models for damage classification and detection, respectively. The proposed method was validated using a dataset reflecting real-world logistics situations, demonstrated its potential to enhance the quality of parcel delivery services.
Abstract
The damage to parcel not only diminishes customer satisfaction with parcel delivery services but also incurs return logistics costs, and this problem can be addressed by detecting damage before parcels reach the customers. Therefore, various deep learning researches related to the parcel damage detection are being conducted. In this study, we introduce a method for parcel box damage detection that not only identifies the presence of damage but also determines its specific type. To achieve this, we employ DCGAN-based data augmentation and use the YOLOv5s-cls and YOLOv5m models for damage classification and detection, respectively. The proposed method was validated using a dataset reflecting real-world logistics situations, demonstrated its potential to enhance the quality of parcel delivery services.
참고문헌 (19)
- "택배이용 종합 안내서" (2020), 국가물류통합정보센터[URL]
- 김미례, "Improvement of Recognition Performance through Refinement of Parcel Damage Classification Algorithm based on CNN", 박사, pp.2022
- 김병관, "수도권 택배 물동량 추정에 관한 탐색적 연구: 경기도를 중심으로", 로지스틱스연구, 31(2), pp.47~58 (2023), 한국로지스틱스학회[DOI]
- 김은강, "딥러닝네트워크를 통한 택배 상자 파손 분류", 한국방송공학회 학술발표대회 논문집, 2022, pp.250~253 (2022)
- "택배 포장하기" (2023), 법제처[URL]
- 양재훈, "택배서비스 이용자가 인지하는 관계효익과 고객충성도에 관한 연구", 로지스틱스연구, 19(3), pp.83~103 (2011), 한국로지스틱스학회[DOI]
- "2023년 7월 온라인쇼핑 동향" (2023), 통계청[URL]
- "택배서비스 만족도 ‘우체국택배’가 가장 높아" (2017), 한국소비자원[URL]
- 한수민, "택배 물류 터미널에서의 화물하차 로봇 도입의 효율성 분석", 로지스틱스연구, 28(6), pp.1~11 (2020), 한국로지스틱스학회[DOI]
- Bochkovskiy, A., "Yolov4 : Optimal speed and accuracy of object detection", arXiv preprint, arXiv:2004.10934, ~ (2020)
- Everingham, M., "The pascal visual object classes(voc)challenge", International journal of computer vision, 88, pp.303~338 (2010)
- Goodfellow, I., "Generative adversarial networks", Communications of the ACM, 63(11), pp.139~144 (2020)
- Lin, T. Y., "Focal loss for dense object detection", 2017 IEEE International Conference on Computer Vision(ICCV), 42(2), pp.2980~2988 (2017)
- Liu, W., "Ssd : Single shot multibox detector", Computer Vision–ECCV 2016, 9905, pp.21~37 (2016)
- Radford, A., "Unsupervised representation learning with deep convolutional generative adversarial networks", arXiv preprint, arXiv:1511.06434, ~ (2015)
- Russakovsky, O., "Imagenet large scale visual recognition challenge", International journal of computer vision, 115, pp.211~252 (2015)
- Salimans, T., "Improved techniques for training gans", NIPS’16: Proceedings of the 30th International Conference on Neural Information Processing Systems, 2016, pp.2234~2242 (2016)
- "yolov5" (2020)[URL]
- Wang, Z., "Multitype damage detection of container using cnn based on transfer learning", Mathematical Problems in Engineering 2021, 2021, pp.1~12 (2021)