Datta / Banerjee Face Detection and Recognition

Theory and Practice
Erscheinungsjahr 2015
ISBN: 978-1-4822-2657-7
Verlag: Taylor & Francis
Format: PDF
Kopierschutz: Adobe DRM (»Systemvoraussetzungen)

E-Book, Englisch, 352 Seiten

Face detection and recognition are the nonintrusive biometrics of choice in many security applications. Examples of their use include border control, driver’s license issuance, law enforcement investigations, and physical access control.

Face Detection and Recognition: Theory and Practice elaborates on and explains the theory and practice of face detection and recognition systems currently in vogue. The book begins with an introduction to the state of the art, offering a general review of the available methods and an indication of future research using cognitive neurophysiology. The text then:

- Explores subspace methods for dimensionality reduction in face image processing, statistical methods applied to face detection, and intelligent face detection methods dominated by the use of artificial neural networks

- Covers face detection with colour and infrared face images, face detection in real time, face detection and recognition using set estimation theory, face recognition using evolutionary algorithms, and face recognition in frequency domain

- Discusses methods for the localization of face landmarks helpful in face recognition, methods of generating synthetic face images using set estimation theory, and databases of face images available for testing and training systems

- Features pictorial descriptions of every algorithm as well as downloadable source code (in MATLAB®/PYTHON) and hardware implementation strategies with code examples

- Demonstrates how frequency domain correlation techniques can be used supplying exhaustive test results

Face Detection and Recognition: Theory and Practice provides students, researchers, and practitioners with a single source for cutting-edge information on the major approaches, algorithms, and technologies used in automated face detection and recognition.

Zielgruppe

Graduates and Researchers in Computer Science, Electronics Engineering, Electrical Engineering, and Instrumentation Engineering.

Autoren/Hrsg.

Datta, Asit Kumar

Datta, Madhura

Banerjee, Pradipta Kumar

Fachgebiete

• Rechtswissenschaften Strafrecht Rechtsmedizin, Forensik
• Technische Wissenschaften Sonstige Technologien | Angewandte Technik Signalverarbeitung, Bildverarbeitung, Scanning
• Mathematik | Informatik EDV | Informatik Informatik Theoretische Informatik

Inhaltsverzeichnis

Introduction

Introduction

Biometric identity authentication techniques

Face as biometric identity

Automated face recognition system

Process flow in face recognition system

Problems of face identification and recognition techniques

Liveness detection for face recognition

Tests and metrics

Cognitive psychology in face recognition

Face detection and recognition techniques

Introduction to face detection

Feature based approaches for face detection

Low level analysis

Active shape model

Feature analysis

Image based approaches for face detection

Statistical approaches

Face recognition methods

Geometric feature based method

Subspace based face recognition

Neural network based face recognition

Correlation based method

Matching pursuit based methods

Support vector machine approach

Selected works on face classifiers

Face reconstruction techniques

Three dimensional face recognition

Subspace based face recognition

Introduction

Principal component analysis

Two dimensional principal component analysis

Kernel principal component analysis

Fisher linear discriminant analysis

Fisher linear discriminant analysis for two class case

Independent component analysis

Face detection by Bayesian approach

Introduction

Bayes decision rule for classification

Gaussian distribution

Bayes theorem

Bayesian decision boundaries and discriminant function

Density estimation using eigenspace decomposition

Bayesian discriminant features (BDF) method for face detection

Modelling of face and non-face pattern

Bayes classification using BDF

Experiments and results

Face detection in colour and infrared images

Introduction

Face detection in colour images

Colour spaces

RGB model

HSI colour model

YCbCr colour space

Face detection from skin regions

Skin modelling

Probabilistic skin detection

Face detection by localizing facial features

Eye map

Mouth map

Face detection in infrared images

Multivariate histogram based image segmentation

Method for finding major clusters from a multivariate histogram

Experiments and results on the colour and IR face image datasets

Utility of facial features

Intelligent face detection

Introduction

Multilayer perceptron model

Learning algorithm

Face detection networks

Training images

Data preparation

Face training

Exhaustive training

Evaluation of face detection for upright faces

Algorithm

Image scanning and face detection

Real time face detection

Introduction

Features

Integral image

Rectangular feature calculation from integral image

ADABOOST

Modified ADABOOST algorithm

Cascade classifier

Face detection using OpenCV

Face space boundary selection for face detection and recognition

Introduction

Face points, face classes and face space boundaries for face detection and recognition

Mathematical preliminaries for set estimation method

Face space boundary selection using set estimation for face detection

Algorithm for global threshold based face detection

Experimental design and result analysis

Face / non-face classification using global threshold during face detection

Comparison between threshold selections by ROC based and set estimation based techniques

Classification of face / non-face regions of an image containing multiple faces

Class specific thresholds of face-class boundaries for face recognition

Experimental design and result analysis on face datasets for face recognition

Description of face dataset

Open test results considering imposters in the system

Recognition rates considering only clients in the system

Evolutionary design for face recognition

Introduction

Genetic algorithms

Implementation

Algorithm

Representation and discrimination

Whitening and rotation transformation

Chromosome representation and genetic operators

The fitness function

The evolutionary pursuit algorithm for face recognition

Frequency domain correlation filters in face recognition

Introduction

PSR calculation

A brief review on correlation filters

Mathematical background of a representative correlation filter

ECPSDF filter design

MACE filter design

MVSDF filter design

Optimal tradeoff (OTF) filter design

Unconstrained correlation filter design

UMACE filter design

OTMACH filter design

Physical requirements in designing correlation filters

Applications of correlation filter in face recognition

Performance analysis of correlation filters in face recognition

Performance evaluation using PSR values

Performance evaluation in terms of %RR and %FAR

Performance evaluation by receiver operating characteristics (ROC) curves

Correlation filters for face detection and recognition in video

Formulation of unconstrained video filter

Mathematical formulation of MUOTSDF

Unconstrained video filter

Distance classifier correlation filter

Application of UVF in video for face detection

Training approach

Testing approach

Face detection in video using UVF

Validation of face detection

Face classification using DCCF

Subspace based face recognition in frequency domain

Introduction

Subspace based correlation filter

Mathematical modelling of 1D subspace based correlation filters

Reconstructed correlation filter using 1D subspace

Optimum projecting image correlation filter using 1D subspace

Face classification and recognition analysis in frequency domain

Test results with 1D subspace analysis

Comparative study in terms of PSRs

Comparative study on %RR and %FAR

Mathematical modelling of 2D subspace based correlation filter

Reconstructed correlation filter using 2D subspace

Test results on 2D subspace analysis

PSR value distribution for authentic and impostors

Comparative performance in terms of %RR

Performance evaluation using ROC analysis

Class specific subspace based nonlinear correlation filter

Formulation of nonlinear correlation filters

Nonlinear optimum projecting image correlation filter

Nonlinear optimum reconstructed image correlation filter

Face recognition analysis using correlation classifiers

Test results

Comparative study on discriminating performances

Comparative performance based on PSR distribution

Performance analysis using ROC

Noise sensitivity

Landmark localization for face recognition

Introduction

Elastic bunch graph matching

Gabor Wavelets

Gabor Jets

The EBGM algorithm

Application to face recognition

Application of frequency domain correlation filter in facial landmark detection

ASEF correlation filter

Formulation of ASEF

Eye detection

Multi correlation approach using landmark filter for facial landmark detection

Design of landmark filter (LF)

Landmark localization with localization filter

Test results

Two dimensional synthetic face generation using set estimation technique

Introduction

Generating face points in the face space taking the features from intra class face images

Face generation using algorithm with intra class features and related peak signal to noise ratio

Generating face points in the face space taking the features from inter class face images

Face generation with inter class features

Rejection of the non-meaningful face and corresponding PSNR test

Generalization capability of set estimation method: generating face images not in training set

Test of significance

Datasets of face images and performance tests for face recognition

Face datasets

ORL dataset

OULU physics dataset

XM2VTS dataset

YALE dataset

Yale-B dataset

MIT dataset

CMU pose, illumination, and expression (PIE) dataset

UMIST dataset

PURDU AR dataset

FERET dataset

Performance evaluation of face recognition algorithms

FERET and XM2VTS protocols

Face recognition grand challenge - FRGC

Face recognition vendor test - FRVT

Multiple biometric grand challenge

Focus of evaluation

Conclusion

Über Autor(innen)

Asit Kumar Datta is a former professor of the University of Calcutta (CU), Kolkata, India, where he served in the Department of Applied Physics and the Department of Applied Optics and Photonics. He holds an M.Tech and Ph.D from the same university. Dr. Datta spent 19 years as a professor and a total of 40 years of teaching and research at the post-graduate level at CU. In addition, he served for 8 years as a principal scientist/principal scientific officer of a CU research center in optical electronics. Widely published in international journals and conference proceedings, Dr. Datta has guided 14 scholars toward their Ph.Ds and has published nearly 125 papers. He has contributed significantly in the areas of photonic computation, photonic and electronic instrumentation, optical communications, and pattern recognition. He represented India at the International Commission on Optics and the International Commission on Illumination.

Madhura Datta is the assistant director of the University Grants Commission-Human Resources Development Center, University of Calcutta, Kolkata, India. She holds an M.Sc in computer and information science, and an M.Tech and Ph.D in computer science and engineering from the University of Calcutta. Her primary areas of research are face detection and recognition. Her work has been featured in various technical publications and conference proceedings, including the Journal of Pattern Recognition Research, Computer Vision and Image Understanding, International Journal of Pattern Recognition and Artificial Intelligence, International Conference on Pattern Recognition and Machine Intelligence, and IEEE International Conference on Intelligent Human Computer Interaction.

Pradipta Kumar Banerjee is an associate professor in the Department of Electrical Engineering of the Future Institute of Engineering and Management, Kolkata, India. He holds a B.Sc, B.Tech, M.Tech, and Ph.D from the University of Calcutta, Kolkata, India. His research interests include computer vision, image processing, frequency domain pattern recognition, biometrics, object detection, recognition, and tracking. He has been published in numerous technical journals, book chapters, and conference proceedings, such as Optics & Laser Technology, Optik - International Journal for Light and Electron Optics, Pattern Recognition Letters, Lecture Notes in Computer Science, 8th International Conference on Advances in Pattern Recognition, IEEE International Conference on Signal and Image Processing, IEEE Conference on Recent Advances in Intelligent Computational Systems, and many more.