基於邊緣運算和LSTM的即時暴力檢測系統

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：33

、訪客IP：18.119.130.36

姓名

賴世晟(Shih-Cheng Lai) 查詢紙本館藏

畢業系所

通訊工程學系

論文名稱

基於邊緣運算和LSTM的即時暴力檢測系統
(Real-time Violence Detection System based on Edge Computing and LSTM)

相關論文

★ UHF頻段RFID彈藥管理系統之設計、實作與評估	★ 移動物偵測與追蹤之IP Camera系統
★ SDN自適應性自動化網路安全之研究	★ Wi-Fi Direct Service 應用於IoT
★ 射頻前端電路應用於載波聚合長期演進技術	★ 3C無線充電裝置運用在車載系統所產生之EMI輻射
★ 基於LoRa技術的物聯網前端防盜警示感測裝置實作與評估	★ DOCSIS 3.1 效能研究與下行通道干擾阻隔之設計
★ 藍芽無線光學投影翻譯筆	★ 手持裝置應用於MIMO ( 8x8 ) Wi-Fi系統之設計
★ 基於無伺服器運算之智慧農業雲端系統設計與研究	★ 嵌入式系統實現電梯物聯網
★ 在802.11 Ad-Hoc網路中基於速率考量之路由協定設計	★ 合作博弈與灰色模糊方法改善無線網路之性能
★ 採用拍賣策略之動態分散式方法於減少叢集小型基地台間干擾之研究	★ 在LTE-A下聚合未授權頻譜及動態分配資源以優化系統效能

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 (2026-7-10以後開放)

摘要(中)

近年來物聯網的應用讓設備之間的聯繫更為緊密，有助於即時傳遞重要
訊息。然而，在現有監控系統中，物聯網的優勢未能充分發揮，許多暴力事
件仍無法及時察覺和干預，導致無法挽回的傷害。這顯示出目前監控系統在
即時檢測和預防暴力事件方面仍有不足。
本論文提出了一種利用樹莓派作為邊緣裝置的智慧監控系統，旨在應對
暴力事件的日益增長及現有監控系統的不足。此系統運用了機器學習技術，
結合移動神經網路（MobileNet）和長短時記憶網路（LSTM），能夠即時檢
測並識別暴力行為，同時向相關單位發送警報，以提高對暴力事件的預防與
即時回應能力。此外，為了兼顧隱私和資料安全，系統採用邊緣運算核心，
這不僅保護了個人隱私，還能高效利用影像資料進行分析。這種設計在增強
公共安全監控效能的同時，也能有效保障個人隱私。

摘要(英)

In recent years, the application of the Internet of Things （IoT） has
significantly tightened the connectivity between devices, facilitating the
instantaneous transmission of crucial information. However, the potential of IoT
has not been fully realized in existing surveillance systems, and many violent
incidents still go undetected and unaddressed in time, resulting in irreparable harm.
This highlights the current shortcomings of surveillance systems in the real-time
detection and prevention of violent events.
This thesis utilizes the Raspberry Pi as an edge device to address the growing
prevalence of violent incidents and the inadequacies of existing monitoring
systems. By employing advanced machine learning technologies, we develop an
intelligent surveillance system. The system integrates Mobile Neural Networks
（MobileNet） and Long Short-Term Memory networks（LSTM） to detect and
identify violent behaviors in real time. Upon identification, it sends alerts to
relevant authorities, enhancing the prevention and immediate response to violent
events. Additionally, to balance privacy and data security, the system employs
edge computing at its core, safeguarding personal privacy while efficiently
analyzing video data. This design not only enhances the effectiveness of public
safety monitoring but also effectively protects individual privacy.

關鍵字(中)

★ 邊緣運算
★ MobileNet
★ LSTM
★ 物聯網

關鍵字(英)

★ Edge Computing
★ MobileNet
★ LSTM
★ IoT

論文目次

摘要 I
Abstract II
誌謝 III
目錄 IV
圖目錄 VII
表目錄 IX
第一章序論 1
1-1 前言 1
1-2 研究動機 2
1-3 論文架構 3
第二章相關研究背景 4
2-1 邊緣運算 4
2-1-1 嵌入式裝置—樹莓派 6
2-1-2 Pi Camera 8
2-2 機器學習 9
2-2-1 卷積類神經網路 CNN 10
2-2-2 長短期記憶 LSTM 13
2-3 資料傳輸 16
2-3-1 SSH 16
2-3-2 SCP 17
2-4 Line 18
2-4-1 Line Notify 18
第三章系統架構與流程 19
3-1 場景分析與架構設計 19
3-1-1 場景分析 19
3-1-2 系統架構 20
3-2 伺服器端 21
3-2-1 收集影像樣本 22
3-2-2 數據預處理 23
3-2-3 機器學習模型架構 25
3-2-4 評估模型與分類閾值 27
3-3 邊緣端 29
3-3-1 即時監控 30
3-3-2 影像預處理 31
3-3-3 模型偵測 32
3-3-4 上傳影像資料及發送Line通知 34
第四章模擬及結果分析 35
4-1 模擬設定 35
4-1-1 機器學習模型設定 35
4-2 結果分析 37
4-2-1 模擬訓練結果 38
4-2-2 模型效能評估 41
4-2-3 實際應用情況 44
第五章結論 47
參考文獻 49

參考文獻

[1] G. Kaur and R. S. Batth, "Edge Computing: Classification, Applications, and Challenges," in 2021 2nd International Conference on Intelligent Engineering and Management (ICIEM), 28-30 April 2021 2021, pp. 254-259, doi: 10.1109/ICIEM51511.2021.9445331.
[2] M. Talebkhah, A. Sali, M. Marjani, M. Gordan, S. J. Hashim, and F. Z. Rokhani, "Edge computing: Architecture, Applications and Future Perspectives," in 2020 IEEE 2nd International Conference on Artificial Intelligence in Engineering and Technology (IICAIET), 26-27 Sept. 2020 2020, pp. 1-6, doi: 10.1109/IICAIET49801.2020.9257824.
[3] A. S. Al-Ahmad and H. Kahtan, "Cloud Computing Review: Features And Issues," in 2018 International Conference on Smart Computing and Electronic Enterprise (ICSCEE), 11-12 July 2018 2018, pp. 1-5, doi: 10.1109/ICSCEE.2018.8538387.
[4] S. Karthikeyan, R. A. Raj, M. V. Cruz, L. Chen, J. L. A. Vishal, and V. S. Rohith, "A Systematic Analysis on Raspberry Pi Prototyping: Uses, Challenges, Benefits, and Drawbacks," IEEE Internet of Things Journal, vol. 10, no. 16, pp. 14397-14417, 2023, doi: 10.1109/JIOT.2023.3262942.
[5] W. Rawat and Z. Wang, "Deep Convolutional Neural Networks for Image Classification: A Comprehensive Review," (in eng), Neural Comput, vol. 29, no. 9, pp. 2352-2449, Sep 2017, doi: 10.1162/NECO_a_00990.
[6] J. Latif, C. Xiao, A. Imran, and S. Tu, "Medical Imaging using Machine Learning and Deep Learning Algorithms: A Review," in 2019 2nd International Conference on Computing, Mathematics and Engineering Technologies (iCoMET), 30-31 Jan. 2019 2019, pp. 1-5, doi: 10.1109/ICOMET.2019.8673502.
[7] C. Junliang, "CNN or RNN: Review and Experimental Comparison on Image Classification," in 2022 IEEE 8th International Conference on Computer and Communications (ICCC), 9-12 Dec. 2022 2022, pp. 1939-1944, doi: 10.1109/ICCC56324.2022.10065984.
[8] A. G. Howard et al., "MobileNets: Efficient Convolutional Neural Networks for Mobile Vision Applications," p. arXiv:1704.04861doi: 10.48550/arXiv.1704.04861.
[9] M. Sandler, A. Howard, M. Zhu, A. Zhmoginov, and L.-C. Chen, "MobileNetV2: Inverted Residuals and Linear Bottlenecks," p. arXiv:1801.04381doi: 10.48550/arXiv.1801.04381.
[10] A. Howard et al., "Searching for MobileNetV3," p. arXiv:1905.02244doi: 10.48550/arXiv.1905.02244.
[11] N. Djeffal, D. Addou, H. Kheddar, and S. A. Selouani, "Noise-Robust Speech Recognition: A Comparative Analysis of LSTM and CNN Approaches," in 2023 2nd International Conference on Electronics, Energy and Measurement (IC2EM), 28-29 Nov. 2023 2023, vol. 1, pp. 1-6, doi: 10.1109/IC2EM59347.2023.10419459.
[12] H. H. Tan and K. H. Lim, "Vanishing Gradient Mitigation with Deep Learning Neural Network Optimization," in 2019 7th International Conference on Smart Computing & Communications (ICSCC), 28-30 June 2019 2019, pp. 1-4, doi: 10.1109/ICSCC.2019.8843652.
[13] T. Ylonen, "SSH Key Management Challenges and Requirements," in 2019 10th IFIP International Conference on New Technologies, Mobility and Security (NTMS), 24-26 June 2019 2019, pp. 1-5, doi: 10.1109/NTMS.2019.8763773.
[14] G. Xu, D. Sha, Y. Xu, and X. Liao, "Dual-Transformer-Based DAB Converter With Wide ZVS Range for Wide Voltage Conversion Gain Application," IEEE Transactions on Industrial Electronics, vol. 65, no. 4, pp. 3306-3316, 2018, doi: 10.1109/TIE.2017.2756601.
[15] M. Cheng, K. Cai, and M. Li, "RWF-2000: An Open Large Scale Video Database for Violence Detection," p. arXiv:1911.05913doi: 10.48550/arXiv.1911.05913.
[16] D. P. Kingma and J. Ba, "Adam: A Method for Stochastic Optimization," p. arXiv:1412.6980doi: 10.48550/arXiv.1412.6980.
[17] 網路資料 on line resources：Raspberry Pi Documentation。取自 https://www.raspberrypi.com/documentation/computers/os.html
[18] 網路資料 on line resources：Raspberry Pi Camera Module 2。取自
https://www.raspberrypi.com/products/camera-module-v2/
[19] 網路資料 on line resources：TensorFlow，MobileNetV3Small。取自
https://www.tensorflow.org/api_docs/python/tf/keras/applications/MobileNe tV3Small
[20] 網路資料 on line resources：Long Short-Term Memory (LSTM)。取自
https://medium.com/@saba99/long-short-term-memory-lstm-fffc5eaebfdc
[21] 網路資料 on line resources：OpenBSD Project，OpenBSD manual page
server，SCP。取自https://man.openbsd.org/scp.1
[22] 網路資料 on line resources：Line。取自
https://zh.wikipedia.org/zh-tw/LINE
[23] 網路資料 on line resources：Line Notify。取自
https://notify-bot.line.me/zh_TW/
[24] 網路資料 on line resources：For Mobile & Edge。取自 https://www.tensorflow.org/lite/guide?hl=zh-tw
[25] 網路資料 on line resources：黃俊毓，以 Google tensorflow 機器學習做影像辨識。取自 https://hackmd.io/@yillkid/ByQ7ySDT8/https%3A%2F%2Fhackmd.io%2F %40yillkid%2FH1GYxKR-t

指導教授

吳中實(Jung-Shyr Wu)

審核日期

2024-7-16

推文