臺灣博碩士論文加值系統

現今視訊監控系統已經廣泛地應用在門禁管制與交通監測等方面。但多數系統僅具備簡易的資料記錄與分析功能，常需專人介入主控監視。本論文主要在設計一個適用於防盜保全且兼具即時(Real-Time)、精確(Accurate)、自動化(Automated)、數位化(Digitized)的安全監控系統。本系統設計主要由一開發板(raspberry pi3)、攝像頭(Webcam c922)及個人電腦所組成，其中以開發板及攝像頭執行Web server即時串流影像，本機電腦執行特徵運算。在人臉辨識方面，我們會先於本機電腦建立人臉資料庫，透過face_recognition模組進行人臉特徵提取，當影像輸入時，系統先利用人臉特徵，擷取出人臉特徵，再以主成分分析的方法，簡化資料，再進行人臉辨識。由於本應用程式是即時系統，如何準確、快速地定位出人臉區域是開發辨識系統的主要目標。

Todays video surveillance systems have been widely used in access control and traffic monitoring. However, most systems only have simple data recording and analysis functions, and often require special personnel to intervene in the main control monitoring. This paper is mainly to design a security monitoring system suitable for anti-theft security and real-time (Real-Time), Accurate, Automated and Digitized. The system design is mainly composed of a development board (raspberry pi3), a camera (Webcam c922) and a personal computer. The development board and the camera execute the web server instant streaming image, and the local computer performs characteristic calculation. In face recognition, we will build a face database in the local computer first, and face feature extraction through the face_recognition module. When the image is input, the system first uses the face feature, extracts the face feature, and then takes the main The method of component analysis, simplifying the data, and then performing face recognition. Since this application is an instant system, how to locate the face area accurately and quickly is the main goal of developing the identification system。

摘要..........................................................i
Abstract......................................................ii
誌謝..........................................................iii
目錄..........................................................iv
表目錄.........................................................v
圖目錄.........................................................vi
第一章 緒論.....................................................1
1.1 研究動機....................................................1
1.2 文獻探討....................................................3
1.3 論文貢獻....................................................3
1.4 論文大綱....................................................4
第二章 相關技術探討..............................................5
2.1 Python程式語言..............................................5
2.2電腦視覺(Computer Vision)....................................6
2.2.1 OpenCV(Open Source Computer Vision Library)..............7
2.2.2 OpenCV模組................................................7
2.3 DLIB介紹....................................................10
2.3.1 face_recognition模組......................................12
2.4 Socket......................................................14
2.4.1主從式架構(Client/Server)...................................15
2.4.2 Socket類型.................................................16
第三章 系統設計...................................................17
3.1系統架構......................................................17
3.2程式架構......................................................18
3.3 Socket連線...................................................19
第四章 系統設計與實作..............................................21
4.1人臉辨識演算法..................................................21
4.1.1方向梯度直方圖 (Histogram of Oriented Gradient,HOG)...........23
4.1.2人臉辨識.....................................................31
4.1.3歐幾里得距離..................................................35
4.2系統實測........................................................38
4.2.1硬體介紹......................................................39
4.2.2系統介紹......................................................41
4.2.3調整容差/靈敏度................................................44
第五章 結論與未來展望................................................47
參考文獻.............................................................49

[1] P. Jothi Thilaga;B. Arshath Khan;A.A. Jones;N. Krishna Kumar "Modern Face Recognition Deep Learning," IEEE Second International Conference on Inventive Communication and Computational Technologies (ICICCT), Coimbatore, India,2018, pp.1947-1951。
[2] Mochamad Vicky Ghani Aziz ; Ary Setijadi Prihatmanto ; Hilwadi Hindersah "Implementation of lane detection algorithm for self-driving car on toll road cipularang using Python language,"IEEE 4th International Conference on Electric Vehicular Technology (ICEVT), Bali, Indonesia,2017。
[3] Yasmeen Shaikh ; V. K. Parvati ; S. R. Biradar"Survey of Smart Healthcare Systems using Internet of Things (IoT) : (Invited Paper),"IEEE International Conference on Communication, Computing and Internet of Things (IC3IoT), Chennai,India, India,2018,pp.508-513。
[4] Sahil Narang;Tarun Nalwa;Tanupriya Choudhury;Nirbhay Kashyap,"An efficient method for security measurement in internet of things,"IEEE International Conference on Communication, Computing and Internet of Things(IC3IoT), Chennai ,India,2018,pp.319-323。
[5] Yeong-Hwa Chang ; Ping-Lun Chung;Hung-Wei Lin"Deep learning for object identification in ROS-based mobile robots,"IEEE International Conference on Applied System Invention (ICASI)2018,Chiba,Japan,pp.66-69。
[6] Amit Chakraborty,"Image Processing and Image Pattern Recognition a Programming Tutorial"IEEE First International Conference on Artificial Intelligence for Industries (AI4I),2018,Laguna Hills, CA, USA, USA,pp.122-123。
[7] Nicolas Diaz Salazar ; Jesus Alfonso Lopez Sotelo ; Gustavo Andres Salazar Gomez,"Application of Transfer Learning for Object Recognition Using Convolutional Neural Networks,"IEEE 1st Colombian Conference on Applications in Computational Intelligence (ColCACI),2018,Medellin, Colombia,pp.1-6。
[8] Bahriye Akay ; Oğuz Kaynar ; Ferhan Demirkoparan,"Deep learning based recommender systems"International Conference on Computer Science and Engineering (UBMK),2017,Antalya,Turkey,pp.645-648。
[9] Erik W. Anderson ; Gilbert A. Preston ; Claudio T. Silva,"Using Python for Signal Processing and Visualization",Computing in Science & Engineering,2010,IEEE,pp.90-95。
[10] Zdena Dobesova,"Programming language Python for data processing,"International Conference on Electrical and Control Engineering,2011,Yichang, China,pp.1866-4869。
[11] 調用攝像頭進行人臉識別，支持多張人臉同時識別
https://github.com/coneypo/Dlib_face_recognition_from_camera
[12] 使用pytorch 構建2D和3D Face對齊
https://github.com/1adrianb/face-alignment
[13] HOG +線性SVM框架的對象檢測器
https://github.com/VladKha/object_detector
[14] Welcome to Python.org
https://www.python.org/
[15] 分佈式可視化搜索和可視化數據分析平台
https://github.com/AKSHAYUBHAT/DeepVideoAnalytics
[16] scikit-learn：Python中的機器學習
https://github.com/scikit-learn/scikit-learn
[17] OpenCV library
https://opencv.org/
[18] Dlib C++ Library
http://dlib.net/
[19] The worlds simplest facial recognition api for Python and the command
line
https://github.com/ageitgey/face_recognition
[20] Python-Socket.PyPi
https://pypi.org/project/python-socketio/