結合第三代行動通訊之嵌入式車用駕駛身份辨識定位系統

An Embedded Telematics System for Driver Identification and Positioning Based on 3G Mobile System

黃銘泉(Ming-Chuan Huang)

指導教授 ： 郭文嘉

元智大學/資訊學院/資訊管理學系/碩士(2009年)

摘要

本論文提出一套車用保全人臉識別系統，主要應用影像處理技術及興趣點偵測，並結合衛星導航系統及行動電話，改善目前被動式防盜之缺失。研究中利用人臉五官特徵之不同，將擷取出的有效影像進行前處理後，結合興趣點偵測及幾何雜湊距離，計算出一組特徵值予以儲存於資料庫中，供後續步驟進行入侵偵測比對。本研究主要分成四個階段來進行，第一階段主要透過影像前處理將所粹取之影像予以正規化，第二階段針對所擷取出來之特徵值個別進行興趣點的粹取，第三階段將所粹取出來之興趣點利用幾何雜湊距離計算出各特徵值並將假設合法者之特徵值存入資料庫中，最後將測試影像予以比對並統計出準確度。本論文所提出的辨識系統可在不同天候不同光源的環境中，皆可於既定的辨識速度下，產生高辨識率的輸出。實驗結果顯示，在430個測試樣本，共4600張測試影像中，本論文提出的車用保全系統平均辨識準確度可達91.33%，可以作為日後車用和家用保全市場發展主動式防盜技術之參考。

關鍵字

人臉辨識 ； 興趣點比對 ； 雜湊距離

並列摘要

In this thesis, we propose an embedded telematics system for driver identification and positioning based on 3G mobile system. Image processing techniques and detection of interest points are used as the basis of recognition for driver identification. The warning messages with GPS information are sent automatically through the 3G mobile phone technologies to overcome the weaknesses of current passive anti-theft technologies. There are four steps in our proposed method. First, the face region are detected and normalized by image preprocessing techniques. Secondly, the interesting points are extracted to represent the facial features. Thirdly, the Hash distance is evaluated according to the extracted interesting points. Finally, the matching results are used to identify the driver’s identity. The identification system proposed in this thesis provides a high-speed and accurate recognition result even in different weather or environments with different lights. There are 430 persons and 4600 images in our simulations. Experimental result shows that the accuracy of the proposed system in this thesis is 91.33%, and it can be used to develop the vehicle-use and household-use active anti-theft technology effectively in the future.

並列關鍵字

Face recognition ； Interest point ； Hash distance

參考文獻

[13]黃登淵，莊國楨，楊晏和，陳南樺，王嘉宏，「複雜背景下多重人臉偵測演算法之研究」，科學與工程技術期刊第三卷 第三期 ，2007

[3]P. N. Belhumeur, J.P. Hespanha, and D.J Kriegmen., “Eigenfaces vs. Fisherfaces: recognition using class specific linear projection,” IEEE Transactions on Pattern Analysis and Machine Intelligence, vo.19 no.7, pp711-720, Jul. 1997.

[4]Roberto Brunelli, Tomaso Poggio., “Face Recognition: Feature versus Templates”, IEEE Transaction on Pattern Analysis and Machine Intelligence, vol.15 no.10, pp.1042-1052, Oct. 1993.

[5]Moghaddam, B. and A. Pentland., “Face Recognition using View-Based and Modular Eigenspaces”, Automatic Systems for the Identification and Inspection of Humans, SPIE vol.2277,pp.12-21, Oct. 1994.

[6]Bernd Heisele, Purdy Ho, Jane Wu, and Tomaso Poggio., “Face recognition: component-based versus global approaches”, Computer Vision and Image Understanding, vol. 91 no.1, pp.6-21, Feb. 2003.