Dissertation > Excellent graduate degree dissertation topics show

Face Recognition Research Using Nonsubsampled Contourlet Transform and Local Binary Pattern

Author: YueXuYao
Tutor: YangHuiXian
School: Xiangtan University
Course: Physical Electronics
Keywords: face recognition Nonsubsampled Contourlet Transform Local Binary Pattern Self-quotient image Fisher algorithm
CLC: TP391.41
Type: Master's thesis
Year: 2013
Downloads: 23
Quote: 0
Read: Download Dissertation

Abstract


Biometric technology has made considerable progress in recent decades, facerecognition technology has been extensively studied because of its special advantages.The identification performance of the face recognition technology in the ideal case isnearing practical use. However, because of the face image vulnerable to the impact offactors such as posture, facial expressions, illumination, and face structure its owncomplexity, face recognition system is still facing many of the key issues that need tobe addressed.Face recognition algorithm is consists of three components: face detection,feature extraction, and identification. In which the face recognition systemperformance is largely relates to feature extraction, so the feature extraction is the keyto face recognition algorithms. In recent years, feature extraction method based onmulti-scale and multi-direction Analysis tools become one of the main facerecognition algorithm and Nonsubsampled Contourlet Transform (NSCT) is a realimage representation method with multi-directional, multi-scale, anisotropy andtranslational invariance. NSCT was introduced to face recognition in this paper, whichis used to extract face features robust to the complex environment, the main work is asfollows:(1) An efficient face recognition method based on NSCT and Local BinaryPattern(LBP) was proposed. Firstly, a face image was decomposed with NSCT, andNSCT coefficients in different scales and various orientations were obtained; LBPoperator was then used to get LBP feature maps by extracting local neighboringrelationship from NSCT coefficients(LNBP); Finally, feature maps were respectivelydivided into several blocks, the concatenated histograms, which were calculated overeach blocks, were used as the face feature (LNBPH). Experimental results usingmulti-channel nearest neighbor classifier based on Euclidean distance show that, theextracted feature using the proposed method is robust to the variation of illumination,expression and pose, and has better recognition performance.(2) Aiming at the large dimension of LNBPH, a new dimension reductionmethod suitable to small sample size and statistical histogram was proposed. Eachelement in the LNBPH can be characterized as a particle, examine each particlediscriminability and extract those particles that can be highly identified as the newface feature. besides, in order to eliminate the impact of illumination on low-frequency sub-band coefficient, Self-quotient image algorithm was used forillumination normalization before extracting the LNBPH feature. The experimentalresults show that, the method can effectively reduce the dimensionality of LNBPHwithout decrease recognition rate in the case of holding LNBPH high recognition rate.

Related Dissertations

  1. Research on Algorithms of 2D Face Template Protection,TP391.41
  2. Research of Video Face Recognition Based on Weighted Voting and Key-Frame Extraction,TP391.41
  3. Face Recognition Method Based on DE,TP391.41
  4. The Research of Image Matching Method Based on Feature Descriptor,TP391.41
  5. Research on Face Recognition Based on Classifer Fusion,TP391.41
  6. Based on Sub-pattern Locality Preserving Projection Approaches Research for Race Recognition,TP391.41
  7. Research of Sparse Represention Based Face Recognition Algorithm,TP391.41
  8. A Novel Approach for Projection by Optimize Local Area in Face Recognition Applications,TP391.41
  9. Research on Deep Structure Learning Algorithms Based on Dynamic Fuzzy Relation,TP181
  10. The Study of 3d Face Recognition System,TP391.41
  11. 3D Face Recognition Based on Biomimetic Pattern Recognition,TP391.41
  12. Research on Robust Manifold Learning Algorithms,TP391.41
  13. The Research for Fractal Feature Extraction and Face Recognition Algorithm,TP391.41
  14. Face Recognition Research Based on Manifold Learning,TP391.41
  15. The Research for Face Recognition Based on Pseudo-Zernike Moment and BP Network,TP391.41
  16. Research on the Key Technique in Face Recognition,TP391.41
  17. Research on Face Recognition Methods Based on Flexible Neural Tree,TP391.41
  18. Pose Adjustment in Face Recognition Based on Range Image,TP391.41
  19. Face recognition based on manifold learning techniques,TP391.41
  20. Facial expression recognition based on sparse representation residuals fusion,TP391.41
  21. Research on Face Recognition Based on AdaBoost Algorithm,TP391.41

CLC: > Industrial Technology > Automation technology,computer technology > Computing technology,computer technology > Computer applications > Information processing (information processing) > Pattern Recognition and devices > Image recognition device
© 2012 www.DissertationTopic.Net  Mobile