生物特征識別已廣泛應用于監(jiān)視應用,法醫(yī)學和刑事調查。由于生物識別系統(tǒng)可以提供比傳統(tǒng)的個人身份驗證系統(tǒng)（例如令牌或密碼）更高的安全性解決方案,其中令牌可能被盜,長密碼或密碼難以記憶并且可能被遺忘。此外,隨著對法醫(yī)學和諸如訪問控制,移民和商業(yè)應用等安全領域的更多安全系統(tǒng)的需求的增加,生物識別系統(tǒng)最近引起了很多關注。近年來,耳印由于其顯著的優(yōu)勢,受到了生物統(tǒng)計學界的廣泛關注。人耳很大并且可以獲得,對年齡和表情穩(wěn)定,并且對于同卵雙胞胎和三胞胎而言也是不同的。隨著使用耳朵生物識別系統(tǒng)作為面部和指紋生物識別系統(tǒng)在許多應用中的興趣日益增加,特別是在監(jiān)視和取證上,其需要壓縮耳朵圖像數據,通過網絡傳輸到特定位置。由于存儲容量和傳輸數據的限制,可能是低質量的無線信道,在應用程序中需考慮圖像壓縮對其系統(tǒng)性能的影響。然而,以前的工作沒有提出壓縮技術對耳朵生物識別系統(tǒng)的影響。因此,本文首先研究和分析已知壓縮算法（JPEG,JPEG2000和BPG）對耳朵識別系統(tǒng)的影響,特別是對兩個公共和可用的耳朵數據庫。最近,JPEG和JPEG2000標準在面部和指紋生物識別等生物識別應用中發(fā)揮了至關重要的作用。因此,這項工作... 

【文章來源】：哈爾濱工業(yè)大學黑龍江省 211工程院校 985工程院校

【文章頁數】：81 頁

【學位級別】：碩士

【文章目錄】：
摘要
Abstract
Chapter1.Context and Contributions
    1.1 Introduction
    1.2 Ear Biometrics
    1.3 Problem Statement and Motivation
    1.4 Thesis Objectives and Contributions
    1.5 Thesis Organization
Chapter2.Related Works for Ear Recognition and the Compression Image Techniques
    2.1 Recent Ear Recognition Approaches
        2.1.1 Geometric Feature Extraction Approaches
        2.1.2 Appearance-Based Feature Extraction Approaches
        2.1.3 Three D-based Feature Extraction Approaches
        2.1.4 Deep Feature Extraction Approaches
        2.1.5 Combining Features Approaches
        2.1.6 Matching Methods for Human Ear Recognition
    2.2 Well-Known Compression Algorithms
        2.2.1 Joint Photographic Experts Group(JPEG)
        2.2.2 Joint Photographic Experts Group2000(JPEG2000)
        2.2.3 Better Portable Graphic(BPG)
Chapter3.The Effects of JPEG and JPEG2000 on Ear Recognition System
    3.1 Introduction
    3.2 The proposed method for compression ear recognition based on JPEG and JPEG
        3.2.1 Ear representation
        3.2.2 JPEG and JPEG2000 for ear image compression
    3.3 Experimental Results
        3.3.1 The effects of JPEG on ear recognition system
        3.3.2 The effects of JPEG2000 on ear recognition system
    3.4 Summary
Chapter4.Deep Features and BPG Compression Algorithm for Ear Recognition
    4.1 Introduction
    4.2 The proposed Approach for compression ear recognition based on deep features and BPG Algorithm
        4.2.1 Better Portable Graphic(BPG)for ear image compression
        4.2.2 Deep Features Extraction
    4.3 Experimental Results
        4.3.1 Datasets and evaluation metrics
        4.3.2 Evaluation various compression techniques on ear deep features
    4.4 Comparison on the different compression techniques on ear recognition
    4.5 Summary
Conclusions
References
List of Publications
Acknowledgements
Resume



本文編號：3405615