Dissertation > Excellent graduate degree dissertation topics show

Infrared Image Face Recognition Using Random Projection and Sparse Representation

Author: LiuBin
Tutor: LiangJiMin
School: Xi'an University of Electronic Science and Technology
Course: Circuits and Systems
Keywords: Face Recognition Visible light Far-infrared light Random projection Sparse Representation
CLC: TP391.41
Type: Master's thesis
Year: 2009
Downloads: 348
Quote: 4
Read: Download Dissertation

Abstract


The face recognition technology has made ??great achievements , but also faces many unresolved problems , the visible light image performance of face recognition system is susceptible to the influence of illumination changes is one of the problems . Thermal infrared images from the illumination changes applied to face recognition can achieve good results . However, due to thermal infrared can not penetrate the glass , if the pending identify a person wearing glasses , the image rendering sunglasses form , will cause a sharp decline of recognition performance . In this paper, the thermal infrared face recognition system glasses cover research to characterize the robustness of classification and recognition method based on random projection and sparse . First map the original high-dimensional image data using random projection to low-dimensional subspace of transformation , to achieve the purpose of data dimensionality reduction . Then using the test sample with a very small number of training samples linear classification and recognition method based on Sparse Representation effective isolation cover noise . For the reconstruction of the test sample is estimated based on various types of characterization factor , this paper proposes the Criterion two classifications are reconstructed the residuals and characterization coefficient effectiveness accumulated factor . The same time , in order to achieve great robustness veiled take images horizontal quartered strategy , random projection and sparse characterization in each decile block respectively , four average criterion value for classification . The database used in this experiment for American the Equinox companies collect and record the visible and far infrared face image database , selected according to need 44 people both wear glasses also do not wear glasses far infrared light and its every frame corresponding to the visible light face image . The CSU face recognition using improved evaluation system of the proposed method with the appearance - based face recognition method , principal component analysis , linear discriminant analysis and Bayesian method , comparative assessment based on the far-infrared and visible light images . The assessment results show that this method can effectively solve the problems of the glasses cover the impact of thermal infrared face image recognition rate , and its recognition performance is better than the principal component analysis , linear discriminant analysis , as well as the Bayesian method .

Related Dissertations

  1. Research on Algorithms of 2D Face Template Protection,TP391.41
  2. Iron, lanthanum -doped nano TiO \u003csub\u003e 2 \u003c/ sub\u003e Preparation and photocatalytic properties,O614.411
  3. Compressive Sensing of the Speech Signal and Its Application in Speech Coding,TN912.3
  4. Research on Face Recognition Methods Based on Flexible Neural Tree,TP391.41
  5. Composite fiber BiVO 4 Preparation and visible light photocatalytic properties of dyeing wastewater,X791
  6. Facial expression recognition based on sparse representation residuals fusion,TP391.41
  7. Wavelets-based Signal Sparse Representation and Its Application in Image Denoising,TP391.41
  8. Doped Modification and Application of Titania Photocatalyst,O643.36
  9. Indoor Wireless Visible Light Communication System Based on Quadrature Amplitude Modulation,TN929.1
  10. Performance Research and Theoretical Design of Visible Light LED Communication System,TN929.1
  11. OOK-OFDM Hybrid Modulation and Channel Coding of Visible Light Communication,TN911.22
  12. The Research on Neural Network with Quadratic Denominator Cubic Rational Spline Function Weight and Its Application,TP183
  13. Based on sparse representation of high spatial resolution remote sensing image texture description Method,TP751
  14. Preparation and Characterization of Graphene/Titanate Nanotubes Photocatalysts,TB383.1
  15. Based on wavelet transform and linear subspace face recognition technology,TP391.41
  16. Based on local binary pattern and Weber local descriptor Face Recognition,TP391.41
  17. Based on sparse representation Face Image Recognition Method,TP391.41
  18. Integration and Application of Clustering Algorithm,TP311.13
  19. Research on Subspace Analysis Based Face Recognition Algorithm,TP391.41
  20. Research on Application of Fractional Differential to Digital Image Processing and Support Vector Machine to Face Recognition,TP391.41
  21. Preparation and Photocatalytic Behavior of Zn2+ Doped TiO2 One-dimensional Nanomaterials,TB383.1

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