English  |  正體中文  |  简体中文  |  Items with full text/Total items : 28611/40649
Visitors : 641313      Online Users : 72
RC Version 4.0 © Powered By DSPACE, MIT. Enhanced by NTU Library IR team.
Scope Adv. Search
LoginUploadHelpAboutAdminister

Please use this identifier to cite or link to this item: http://ntour.ntou.edu.tw:8080/ir/handle/987654321/54120

Title: 基於深度學習之即時人臉辨識實作
Implementation of a Deep-Learning Based Real-Time Face Recognition System
Authors: Yu, Shih-Huai
余世懷
Contributors: NTOU:Department of Electrical Engineering
國立臺灣海洋大學:電機工程學系
Keywords: 深度學習;網路攝影機;Raspberry Pi 3B+;Intel Movidius;卷積神經網路;Opencv;Tensorflow
deep learning;webcam;Raspberry Pi 3B+;Intel Movidius neural computing stick;convolutional neural network;Opencv;Tensorflow
Date: 2019
Issue Date: 2020-07-09T03:02:09Z
Abstract: 在現今全球社會的發展當中,科技為人們帶來了非常大的便利性,在以前的時代中,電腦要辨識出一個物體大多是以形狀和輪廓辨識,若要完整確認物體的本質是非常困難的。自從深度學習演算法出現後,更可以明確的定義出物體的本質是什麼,其所辯識到的種類也更加多元化。 本論文是以Webcam結合Raspberry Pi 3B+以及Intel Movidius進而達成人臉辨識之實作,運用到的深度學習方法是以卷積神經網路下去做實現,系統的運作是以Raspberry Pi 3B+開發板當作基底並連接上網路攝影機以及完成Intel Movidius套件的安裝步驟,之後在電腦上以Opencv及Tensorflow深度學習框架訓練完人臉辨識並生成權重檔,之後把訓練完成的權重檔載入Raspberry Pi 3B+上並配合Intel Movidius負責監控及展示出人臉辨識結果。本論文成果可以應用在一些需要管理人員進出的辦公室以達到即時的管理監控。
Science and technology have brought great convenience to the development of the global human society nowadays. In the past, computers often recognized an object by its shape and contour. It was very difficult to completely confirm the nature of the object. Since the advance of the deep learning algorithm, the ability of recognizing the nature of the object is greatly improved, and the types of objects that can be identified become more diverse. The goal of this thesis is to develop and implement a deep-learning based real-time face recognition system. It utilizes a webcam combined with a Raspberry Pi 3B+ and an Intel Movidius neural computing stick to achieve the task of real-time face recognition. The deep learning method used is the convolutional neural network. The operation of the system is by using the Raspberry Pi development board as the base to connect to the network camera and to complete the installation steps of the Intel Movidius neural computing stick. The face recognition is then completed on the computer with the Opencv and Tensorflow deep learning framework to generate the weight file. The weight file is then given to the Intel Movidius neural computing stick and the Raspberry Pi 3B+ to calculate and display the face recognition result. The results of this thesis can be applied to offices or communities that require entrance management to reduce labor costs.
URI: http://ethesys.lib.ntou.edu.tw/cgi-bin/gs32/gsweb.cgi?o=dstdcdr&s=G0010653050.id
http://ntour.ntou.edu.tw:8080/ir/handle/987654321/54120
Appears in Collections:[電機工程學系] 博碩士論文

Files in This Item:

File SizeFormat
index.html0KbHTML31View/Open


All items in NTOUR are protected by copyright, with all rights reserved.

 


著作權政策宣告: 本網站之內容為國立臺灣海洋大學所收錄之機構典藏,無償提供學術研究與公眾教育等公益性使用,請合理使用本網站之內容,以尊重著作權人之權益。
網站維護: 海大圖資處 圖書系統組
DSpace Software Copyright © 2002-2004  MIT &  Hewlett-Packard  /   Enhanced by   NTU Library IR team Copyright ©   - Feedback