博碩士論文 111852008 詳細資訊




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:110 、訪客IP:3.148.117.237
姓名 蘇育霆(SU, YU-TING)  查詢紙本館藏   畢業系所 生醫科學與工程學系
論文名稱 AIoT邊緣運算即時領餐人流計數系統
(AIoT Edge Computing Real-Time Dining Hall Traffic Counting System)
相關論文
★ 使用滾球篩選睡眠紡錘波檢測★ 利用深度學習產生去骨電腦斷層掃描血管造影改進椎動脈分割
★ 評估深度卷積神經網路用於檢測和分割Chest X-ray圖像中的鎖骨骨折★ 自然語言處理於病例情感分析分類器及句子相似度計算
★ 以圓柱採樣訓練深度神經網絡改進頭頸部電腦斷層掃描的骨骼偵測和分割★ 使用深度學習模型自動分割黑血磁共振腦血管管壁
★ 利用自然語言處理在胸腔X-Ray的自由文本病歷報告中標記識別心臟肥大的檔案和句子★ 肺炎診斷導向之深度學習電腦斷層掃瞄影像分割
★ 基於角色情感互動與主動式照護的生成式AI模型能力研究★ 使用SpaCy NER標記胸部放射檢查報告:與 CheXpert Labeler 的比較
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]   [檢視]  [下載]
  1. 本電子論文使用權限為同意立即開放。
  2. 已達開放權限電子全文僅授權使用者為學術研究之目的,進行個人非營利性質之檢索、閱讀、列印。
  3. 請遵守中華民國著作權法之相關規定,切勿任意重製、散佈、改作、轉貼、播送,以免觸法。

摘要(中) 「人流」在各領域皆具有重要性與應用價值。在生產管理方面,透過人流可以評估生產線效率、發現瓶頸、優化布局、提高生產效率和品質。在行銷管理中,人流可用於評估顧客流量、消費者行為,指導陳列和促銷策略,提高銷售額和客戶滿意度。在人力資源管理中,它有助於預測人力需求,調整員工配置,改善工作環境,提高員工滿意度和工作效率。在財務管理方面,人流用於評估營業時間的利用率,控制成本,提高經營效率,增加收入。總之,人流成為企業提升生產效率、行銷策略、人力資源配置和財務管理的關鍵工具,促進企業永續發展。
員工餐廳在食材、成本和供應商管理方面遇到困難。難以準確預測麵食類和飯菜類的需求量,可能導致食材浪費或供應不足,增加成本。同時,即時領餐人流可以提供當前的備菜量是否足夠,是否需要增加解決可能導致庫存積壓和供餐品質問題,影響供應商關係。缺乏即時人流計數也將影響與供應商的合作,難以精確反映需求,可能導致訂單錯誤和庫存管理問題。
這次研究利用CSI Camera捕捉影像,傳送至Jetson Orin Nano進行即時影像處理,搭配YOLOv8進行人物檢測及追蹤,再透過RESTful API將結果以web API方式提供其他系統取得人數數據,實現高效即時人流計數。
人流計數系統可以滿足營養科在領餐人流計數,不過在這次的研究中發現需要有環境控制上的配合才可以逹到更精準的人流計數,在研究過程中因YOLOv8的人物偵測及追蹤無法滿足計數人數的需求,所以我加上了使用基準線的方式來計數人流,在實際上線後發現需要有環境上的控制才可以逹到精準的人數計數功能,在人流計數系統上的數字也可以讓醫院的儀表板取得當下的領餐人數,營養科同仁可以依據歷史數據評估當前用餐人數是否合理,進而評估目前的備菜量是否足夠、是否需做調整,在用餐成本支出部分,因為有了麵食類及飯菜類的領餐人數,所以當成本支出異常時可以協助快速分辨是麵食類或飯菜類領用異常,不在像以往只有指紋機的總額數據。
摘要(英) "Foot traffic" holds significant importance and application value across various fields. In production management, analyzing foot traffic helps assess production line efficiency, identify bottlenecks, optimize layouts, and enhance both productivity and quality. In marketing management, it is used to evaluate customer flow and consumer behavior, guide merchandise display and promotional strategies, thereby increasing sales and customer satisfaction. In human resources management, it aids in forecasting manpower needs, adjusting staff allocation, improving work environments, and boosting employee satisfaction and work efficiency. In financial management, foot traffic is used to evaluate the utilization of business hours, control costs, improve operational efficiency, and increase revenue. Overall, foot traffic has become a key tool for businesses to enhance production efficiency, marketing strategies, human resource allocation, and financial management, promoting sustainable development.
The staff cafeteria faces difficulties in managing ingredients, costs, and suppliers. The challenge of accurately predicting the demand for noodle and rice dishes can lead to either wastage of ingredients or insufficient supply, increasing costs. Real-time foot traffic data during meal pickup can indicate whether the current preparation is sufficient or if adjustments are needed to address potential stockpile and meal quality issues, impacting supplier relationships. A lack of real-time foot traffic counting also affects cooperation with suppliers, as it is difficult to accurately reflect demand, potentially leading to order errors and inventory management issues.
This study employs a CSI Camera to capture images, which are then processed in real time by a Jetson Orin Nano, combined with YOLOv8 for person detection and tracking. The results are then made available through a RESTful API via a web API method for other systems to access the foot traffic data, achieving efficient and real-time foot traffic counting.
The foot traffic counting system can satisfy the needs of the nutrition department in counting meal pickups. However, this study found that environmental control is necessary to achieve more accurate foot traffic counts. Due to the inadequacy of YOLOv8′s person detection and tracking for counting, I incorporated a baseline method to count foot traffic. Upon deployment, it was found that environmental control is necessary to achieve precise foot traffic counting. The numbers from the foot traffic counting system can also provide the hospital dashboard with current meal pickup counts, allowing nutrition department colleagues to use historical data to evaluate whether current dining numbers are reasonable, assess if the current food preparation is sufficient, and make necessary adjustments. In terms of dining cost expenditure, having the counts for noodle and rice dish pickups helps quickly identify if there are any anomalies in the costs of these specific categories, unlike in the past when only total data from fingerprint machines was available.
關鍵字(中) ★ AIoT
★ Edge Computing
★ people counting
關鍵字(英)
論文目次 中文摘要 i
英文摘要 iii
致謝 v
目錄 vi
圖片目錄 viii
一、 緒論 1
1-1 研究動機與背景 1
1-2 人工智慧物聯網 2
1-3 邊緣運算(Edge Computing) 2
1-4 AI模型與邊緣運算整合 3
1-5 AIoT架構的應用與益處 3
1-6 文獻探討 4
二、 YOLOv8 Multi-Object tracking (YOLOv8多物件追蹤) 5
2-1 用於 YOLOv8 物件追蹤與 OpenCV整合 6
2-2 什麼是物件追蹤 6
三、 RESTful API概論 7
四、 CSI 相機(Camera Serial Interface Camera) 10
五、 研究內容與方法 10
5-1 研究目的 10
5-2 人流計數場域 11
5-3 系統架構 13
5-4 系統硬體設備環境 14
5-5 系統環境建置 17
5-6 人流計數功能 27
5-7 數據串接 36
5-8 計數控制管理介面 39
5-9 數據擷取 40
六、 結果與分析 43
七、 結論 52
參考文獻 54
參考文獻 [1] S. M. SALAHUDDIN MORSALIN。「影像邊緣運算應用之人工智慧網絡設計與實現」。博士論文,國立雲林科技大學電子工程系,2023。<https://hdl.handle.net/11296/85kpyn>。

[2] Raj, Emmanuel. "Edge MLOps framework for AIoT applications." (2020).

[3] Cao, Keyan, et al. "An overview on edge computing research." IEEE access 8 (2020): 85714-85728.

[4] Multi-Object Tracking with Ultralytics YOLO,取自https://docs.ultralytics.com/modes/track/

[5] Terven, Juan, Diana-Margarita Córdova-Esparza, and Julio-Alejandro Romero-González. "A comprehensive review of yolo architectures in computer vision: From yolov1 to yolov8 and yolo-nas." Machine Learning and Knowledge Extraction 5.4 (2023): 1680-1716.

[6] Kiran, Siripuri, and Shoban Babu Sriramoju. "A Study on the Applications of IOT." Indian Journal of Public Health Research & Development 9.11 (2018): 1173-1175.

[7] Simić, Miloš, et al. "Edge computing system for large-scale distributed sensing systems." ICIST (2018): 36-39.

[8] Sun, Shijie, et al. "Benchmark data and method for real-time people counting in cluttered scenes using depth sensors." IEEE Transactions on Intelligent Transportation Systems 20.10 (2019): 3599-3612.

[9] Wetzel, Johannes, Astrid Laubenheimer, and Michael Heizmann. "Joint probabilistic people detection in overlapping depth images." IEEE access 8 (2020): 28349-28359.

[10] Depatla, Saandeep, and Yasamin Mostofi. "Crowd counting through walls using WiFi." 2018 IEEE international conference on pervasive computing and communications (PerCom). IEEE, 2018.

[11] Ibrahim, Osama Talaat, Walid Gomaa, and Moustafa Youssef. "CrossCount: A deep learning system for device-free human counting using WiFi." IEEE Sensors Journal 19.21 (2019): 9921-9928.

[12] Madhira, Kartik, and Aditya Shukla. "Pedestrian flow counter using image processing." 2017 International Conference on Energy, Communication, Data Analytics and Soft Computing (ICECDS). IEEE, 2017.

[13] Wang, Yameng, et al. "Counting people with support vector regression." 2014 10th International Conference on Natural Computation (ICNC). IEEE, 2014.

[14] Li, Shengye, et al. "Supervised people counting using an overhead fisheye camera." 2019 16th IEEE international conference on advanced video and signal based surveillance (AVSS). IEEE, 2019.

[15] Duan, Zhihao, et al. "Rapid: rotation-aware people detection in overhead fisheye images." Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops. 2020.

[16] Li, Guangqin, et al. "Real-time top-view people counting based on a kinect and nvidia jetson tk1 integrated platform." 2016 IEEE 16th international conference on data mining workshops (ICDMW). IEEE, 2016.

[17] Le, Manh Cuong, My-Ha Le, and Minh-Thien Duong. "Vision-based people counting for attendance monitoring system." 2020 5th International Conference on Green Technology and Sustainable Development (GTSD). IEEE, 2020.

[18] Guo, Qiang, et al. "A fast occluded passenger detector based on MetroNet and Tiny MetroNet." Information Sciences 534 (2020): 16-26.

[19] Tyndall, Ash, Rachel Cardell-Oliver, and Adrian Keating. "Occupancy estimation using a low-pixel count thermal imager." IEEE Sensors Journal 16.10 (2016): 3784-3791.

[20] Hagenaars, Erik, et al. "Single-pixel thermopile infrared sensing for people counting." IEEE Sensors Journal 21.4 (2020): 4866-4873.

[21] Gorelick, Micha, and Ian Ozsvald. High Performance Python: Practical Performant Programming for Humans. O′Reilly Media, 2020.

[22] Yen, Mao-Hsu, et al. "Adaptive Indoor People-Counting System Based on Edge AI Computing." IEEE Transactions on Emerging Topics in Computational Intelligence (2023).

[23] Zhang, Yifu, et al. "Bytetrack: Multi-object tracking by associating every detection box." European conference on computer vision. Cham: Springer Nature Switzerland, 2022.

[24] Yu, Fengwei, et al. "Poi: Multiple object tracking with high performance detection and appearance feature." Computer Vision–ECCV 2016 Workshops: Amsterdam, The Netherlands, October 8-10 and 15-16, 2016, Proceedings, Part II 14. Springer International Publishing, 2016.

[25] Xie, Guobo, and Wen Lu. "Image edge detection based on opencv." International Journal of Electronics and Electrical Engineering 1.2 (2013): 104-106.

[26] Soleimanitaleb, Z., and M. A. Keyvanrad. "Single object tracking: A survey of methods, datasets, and evaluation metrics. arXiv 2022." arXiv preprint arXiv:2201.13066.

[27] YOLOv8 Object Tracking and Counting with OpenCV,取自https://learnopencv.com/yolov8-object-tracking-and-counting-with-opencv/

[28] Hagenaars, Erik, et al. "Single-pixel thermopile infrared sensing for people counting." IEEE Sensors Journal 21.4 (2020): 4866-4873.

[29] Fielding, Roy Thomas. Architectural styles and the design of network-based software architectures. University of California, Irvine, 2000.

[30] Chen, Shaobin, and Wei Lin. "Embedded system real-time vehicle detection based on improved YOLO network." 2019 IEEE 3rd advanced information management, communicates, electronic and automation control conference (IMCEC). IEEE, 2019.

[31] Luu, P. V., et al. "An AI-based web surveillance system using raspberry Pi." Journal of Advances in Technology and Engineering Research 5.6 (2019): 231-242.

[32] NVIDIA JetPack Documentation,取自https://docs.nvidia.com/jetson/jetpack/install-jetpack/index.html

[33] https://blog.cavedu.com/2022/03/26/nvidia-jetson-tx2-nx-setup-ssd/

[34]https://docs.nvidia.com/jetson/archives/r36.2/DeveloperGuide/IN/QuickStart.html

[35] Lubanovic, Bill. Introducing Python: Modern Computing in Simple Packages. " O′Reilly Media, Inc.", 2014.
指導教授 黃輝揚(Huang, Hui-Yang) 審核日期 2024-7-11
推文 facebook   plurk   twitter   funp   google   live   udn   HD   myshare   reddit   netvibes   friend   youpush   delicious   baidu   
網路書籤 Google bookmarks   del.icio.us   hemidemi   myshare   

若有論文相關問題,請聯絡國立中央大學圖書館推廣服務組 TEL:(03)422-7151轉57407,或E-mail聯絡  - 隱私權政策聲明