智慧製造標籤系統：基於YOLO模型與圖片轉編碼技術之整合應用

以作者查詢圖書館館藏

、以作者查詢臺灣博碩士

、以作者查詢全國書目

、勘誤回報

、線上人數：25

、訪客IP：18.97.14.91

姓名

廖唯荏(Wei-Jen Liao) 查詢紙本館藏

畢業系所

資訊管理學系在職專班

論文名稱

智慧製造標籤系統：基於YOLO模型與圖片轉編碼技術之整合應用
(Smart Manufacturing Label System: An Integrated Application of YOLO Model and Image-to-Code Techniques)

相關論文

★ 以機器學習技術為基礎建構新生兒孕育健康狀態預測模型	★ 電子病歷縮寫消歧與一對多分類任務
★ 使用文字探勘與深度學習技術建置中風後肺炎之預測模型	★ 精準社群廣告投資策略：以機器學習技術為基礎之社會影響力管理模式
★ 智慧活躍老化之實現：以資料驅動為基礎之AI長者在地交友推薦模式	★ 整合深度學習技術與SOR理論之資訊情緒傳遞性探索：新聞生成特質與資訊情感傳播行為
★ 基於混合過濾的電影推薦系統	★ 以 Reddit 使用者生成內容探討糖尿病照護社會支持
★ 防範於未然：基於機器學習技術之網路入侵偵測系統	★ 可靠度驗證實驗室導入人工智慧技術的可行性探討 -以A公司為例
★ 智慧共同照護之實現: 以資料驅動為基礎之 AI 糖尿病個案管理模式	★ Non-Touch Cooperation: An Interactive Mechanism Design Based on Mid-Air Gestures
★ 基於 UGC 的反脆弱社交口碑策略:以臺灣飯店業為例	★ 基於機器學習技術之誘導式評論過濾機制：以餐廳評論為例
★ 資料驅動的球隊經營：NBA球隊競爭力與抱團策略之研究	★ 基於顧客評論觀點揭示組織調整策略和顧客體驗優化：以 Covid-19 餐飲業為例

檔案

[Endnote RIS 格式]

[Bibtex 格式]

[相關文章]

[文章引用]

[完整記錄]

[館藏目錄]

至系統瀏覽論文 (2030-7-11以後開放)

摘要(中)

工業4.0時代的來臨，智慧製造的數位轉型趨勢已成為製造業提升競爭力的關鍵策略。在全球製造環境日益複雜且快速變化的背景下，標籤系統作為產品識別、追蹤與供應鏈管理的基礎，其效率與靈活性直接影響製造流程的整體表現。然而，傳統標籤設計與管理方法存在顯著缺陷：過度依賴人工編碼導致效率低下、專業軟體的高成本與有限授權造成資源瓶頸、標籤模板難以快速適應多樣化的客製需求，以及跨語系與國際標準的複雜要求增加了設計難度。這些問題不僅延長了產品上市時間，也增加了製造成本，成為製造業數位轉型過程中面臨的重大挑戰之一。傳統標籤設計依賴人工編碼與專業軟體，不僅耗時費力，每月人力成本高達10萬新台幣，且受限於軟體授權與專業人才的可用性。因此，本研究基於以上製造業產業現況所面臨的挑戰，提出一套整合YOLO深度學習模型與圖片轉程式碼技術的智慧製造標籤系統。本研究首先系統性比較YOLOv7、YOLOv8與YOLOv9三種模型在標籤辨識任務的表現，找出最適合的影像辨識模型；其次，開發圖片轉程式碼技術，實現從標籤視覺元素到程式碼的自動轉換；最後，整合兩項技術為標籤自動化系統。研究結果顯示，表現最佳的YOLO v9模型在標籤元素辨識上達到88.9%的準確率，而本研究提出之智慧製造標籤系統將標籤設計時間從平均2-8小時縮減至10分鐘，有效降低了約97%的設計成本。本研究智慧製造標籤系統的開發有效以資料驅動為基礎的深度學習技術實現數位轉型，透過程式碼自動化生成的流程最佳化以及透過數位化技術降低對實體資源依賴的虛實整合。透過智慧製造標籤系統的實際應用，不僅為製造業解決了傳統方法的效率與成本問題，亦可謂組織本身帶來多重的策略優勢，如：實現標籤設計的民主化，使非專業人員也能創造高品質標籤；提高製造靈活性，使企業更快速響應市場變化；最佳化資源配置，讓工程師更專注於具專業性與創新性的任務；提升標籤一致性與準確性，減少因人為錯誤導致的品質問題；同時促進製造資料的標準化與整合，為後續的大數據分析與智慧決策奠定基礎。本研究不僅為製造業標籤管理提供了實用的自動化解決方案，也為圖片轉程式碼技術在製造業的應用提供了個案實證，更為製造業提供明確的數位轉型可複製技術路徑指引。

摘要(英)

The emergence of Industry 4.0 has positioned digital transformation in smart manufacturing as a vital strategy for manufacturing competitiveness. In today′s complex and dynamic global manufacturing landscape, label systems underpin product identification, tracking, and supply chain management, with their efficiency directly affecting overall manufacturing performance. Traditional label design and management methods face several limitations: dependency on manual coding reduces efficiency, expensive specialized software with limited licenses creates bottlenecks, templates fail to adapt to diverse customization needs, and multilingual requirements complicate design processes. These issues extend time-to-market, increase costs, and present significant barriers to manufacturing digital transformation. Current approaches rely heavily on manual processes that cost approximately NT$100,000 monthly and are constrained by both software availability and specialized skill requirements. This study addresses these industrial challenges by developing a smart manufacturing label system that integrates YOLO deep learning models with image-to-code technology. The study compares YOLO v7, YOLO v8, and YOLO v9 performance in label recognition tasks, develops technology for converting visual elements to program code, and combines these technologies into an automated system. Findings indicate that the most effective YOLO variant achieved 88.9% accuracy in element recognition, while the integrated system reduced design time from 2-8 hours to 10 minutes, decreasing design costs by approximately 97%. This system implementation advances digital transformation through data-driven deep learning techniques, optimizes processes through automated code generation, and reduces dependency on physical resources through digital integration. The practical applications provide numerous strategic advantages: enabling non-specialists to create high-quality labels, increasing manufacturing responsiveness, redirecting engineering talent to more innovative tasks, improving label consistency, and standardizing manufacturing data to support advanced analytics. This research offers both a practical solution for label management and empirical evidence for image-to-code applications in manufacturing, ultimately providing a reproducible pathway for digital transformation in manufacturing operations.

關鍵字(中)

★ 智慧製造
★ YOLO
★ 圖片轉程式碼
★ 標籤系統
★ 深度學習
★ 數位轉型

關鍵字(英)

★ Smart Manufacturing
★ YOLO
★ Image-to-Code
★ Label System
★ Deep Learning
★ Digital Transformation

論文目次

目錄
摘要 i
Abstract iii
誌謝 v
目錄 vi
圖目錄 ix
表目錄 x
第一章緒論 1
1-1 研究背景與動機 1
1-2 研究目的 3
1-3 研究範圍與限制 4
1-4 論文架構 6
第二章文獻探討 8
2-1 智慧製造與標籤管理系統 8
2-2 YOLO深度學習模型 10
2-3 圖片轉程式碼技術 (Image-to-Code, ITC) 11
2-4 製造業標籤管理 13
第三章研究方法 16
3-1 系統流程 16
3-2 研究設計與方法 17
3-3 資料來源 19
3-4 資料前處理 20
3-4-1 資料轉換 20
3-4-2 資料清洗 22
3-4-3 ZPL轉JPG 23
3-4-4 特徵標註 24
3-5 YOLO 模型比較實驗 25
3-6 圖片轉程式碼技術開發 28
3-7 系統整合與評估 29
第四章研究結果與分析 31
4-1 模型訓練結果 31
4-1-1 YOLOv7 模型訓練結果 31
4-1-2 YOLOv8模型訓練結果 33
4-1-3 YOLOv9模型訓練結果 35
4-1-4 結果比較 37
4-2 程式碼轉換 40
4-2-1程式碼流程 40
4-2-2 編碼邏輯 41
4-3 效率驗證結果 43
第五章結論 45
5-1 結論 45
5-2研究限制 46
5-3未來研究與建議 47
參考文獻 48

參考文獻

經濟部（2024）。進出口貿易統計。取自經濟部網站：
https://www.moea.gov.tw/Mns/populace/news/News.aspx?kind=1&menu_id=40&nnew_id=113561
經濟部（2024）。製造業單位產出勞動成本指數。取自經濟部網站：https://service.moea.gov.tw/EE521/common/Common.aspx?code=J&no=6
汪建南、馬雲龍 (2016) 。工業4.0的國際發展趨勢與臺灣因應之道。國際金融參考資料，69 2016.12[民105.12] ， 133-155。
MOPS公開資訊觀測站（2024）。https://mops.twse.com.tw/mops/#/web/home
Xu Y, Bo L, Sun X, Li B, Jiang J, Zhou W (2021) image2emmet: Automatic code generation from web user interface image. J Softw Evol Process 33(11).
Grieves, M. (2014). Digital Twin: Manufacturing Excellence through Virtual Factory Replication. [White Paper, Florida Institute of Technology].
Beltramelli, T. (2018). pix2code: Generating code from a graphical user interface screenshot. Proceedings of the ACM SIGCHI symposium on engineering interactive computing systems,
Bi, H., Wen, V., & Xu, Z. (2023). Comparing one-stage and two-stage learning strategy in object detection. Appl. Comput. Eng, 5(1), 171-177.
Chen, W., Chen, L., & Wu, Y. (2024). An efficient and effective transformer decoder-based framework for multi-task visual grounding. European Conference on Computer Vision,
Fernández-Caramés, T. M., & Fraga-Lamas, P. (2018). A review on human-centered IoT-connected smart labels for the industry 4.0. IEEE access, 6, 25939-25957.
Hagen, H., & Suchet, C. (2020). Small labels, big challenges: solutions for advanced therapy labeling. Cell Gene Ther. Insights, 6(8), 1183-1195.
Hindarto, D. (2023). Exploring YOLOv8 Pretrain for Real-Time Detection of Indonesian Native Fish Species. Sinkron: jurnal dan penelitian teknik informatika, 7(4), 2776-2785.
Imran, A., Hulikal, M. S., & Gardi, H. A. (2024). Real Time American Sign Language Detection Using Yolo-v9. arXiv preprint arXiv:2407.17950.
Johnson, K., & Smith, N. (2022). Industrial 2040: A Vision of the Future State of Design and Manufacturing.
Kolthoff, K., Kretzer, F., Fiebig, L., Bartelt, C., Maedche, A., & Ponzetto, S. P. (2024). Zero-Shot Prompting Approaches for LLM-based Graphical User Interface Generation. arXiv preprint arXiv:2412.11328.
Koschmider, A., Figl, K., & Schoknecht, A. (2016). A comprehensive overview of visual design of process model element labels. Business Process Management Workshops: BPM 2015, 13th International Workshops, Innsbruck, Austria, August 31–September 3, 2015, Revised Papers 13,
Redmon, J., Divvala, S., Girshick, R., & Farhadi, A. (2016). You only look once: Unified, real-time object detection. Proceedings of the IEEE conference on computer vision and pattern recognition,
Sheng, V. S., Provost, F., & Ipeirotis, P. G. (2008). Get another label? improving data quality and data mining using multiple, noisy labelers. Proceedings of the 14th ACM SIGKDD international conference on Knowledge discovery and data mining,
Silva‐Rodríguez, V., Nava‐Muñoz, S. E., Castro, L. A., Martínez‐Pérez, F. E., Pérez‐González, H. G., & Torres‐Reyes, F. (2020). Classifying design‐level requirements using machine learning for a recommender of interaction design patterns. IET Software, 14(5), 544-552.
Sundaram, S., & Zeid, A. (2023). Artificial intelligence-based smart quality inspection for manufacturing. Micromachines, 14(3), 570.
Wang, Y. (2021). Research on handwritten note recognition in digital music classroom based on deep learning. Journal of internet technology, 22(6), 1443-1455.
Wang, Y., Wang, H., & Xin, Z. (2022). Efficient detection model of steel strip surface defects based on YOLO-V7. IEEE access, 10, 133936-133944.
Xiao, L., Wang, S., & Jin, D. (2023). Towards Intelligent Manufacturing: An AI-Based Machine Vision Detection System. 2023 China Automation Congress (CAC),
Zhang, C., Chen, Y., Chen, H., & Chong, D. (2024). Industry 4.0 and its implementation: A review. Information Systems Frontiers, 26(5), 1773-1783.
Zhang, X., De Greef, L., Swearngin, A., White, S., Murray, K., Yu, L., Shan, Q., Nichols, J., Wu, J., & Fleizach, C. (2021). Screen recognition: Creating accessibility metadata for mobile applications from pixels. Proceedings of the 2021 CHI Conference on Human Factors in Computing Systems,
Zhao, J., Yan, K., Zhao, Y., Guo, X., Huang, F., & Li, J. (2021). Transformer-based dual relation graph for multi-label image recognition. Proceedings of the IEEE/CVF international conference on computer vision,
Zion Market Research. (2023, June). Smart Bullets Market Size, Share, Growth Report & Trends by Technology, 2023-2030. Zion Market Research.
Ultralytics. (2023). YOLOv8: Real-time Object Detection. Retrieved from https://github.com/ultralytics/ultralytics

指導教授

曾筱珽(Hsiao-Ting,Tseng)

審核日期

2025-7-8

推文