Recognition of Face Biometrics

Recognition of Face Biometrics

Pooja Sharma
Copyright: © 2018 |Pages: 18
DOI: 10.4018/978-1-5225-2848-7.ch017
OnDemand:
(Individual Chapters)
Available
$37.50
Current Special Offers
No Current Special Offers
TOTAL SAVINGS: $37.50

Abstract

In the proposed chapter, a novel, effective, and efficient approach to face recognition is presented. It is a fusion of both global and local features of images, which significantly achieves higher recognition. Initially, the global features of images are determined using polar cosine transforms (PCTs), which exhibit very less computation complexity as compared to other global feature extractors. For local features, the rotation invariant local ternary patterns are used rather than using the existing ones, which help improving the recognition rate and are in alignment with the rotation invariant property of PCTs. The fusion of both acquired global and local features is performed by mapping their features into a common domain. Finally, the proposed hybrid approach provides a robust feature set for face recognition. The experiments are performed on benchmark face databases, representing various expressions of facial images. The results of extensive set of experiments reveal the supremacy of the proposed method over other approaches in terms of efficiency and recognition results.
Chapter Preview
Top

Literature Survey

Global Descriptors

Various global descriptors include Hu’s (1962) seven moment invariants based on geometric moments, the three orthogonal rotation invariant moments (ORIMs) viz. Zernike moments (ZMs) (Teague, 1980), pseudo Zernike moments (PZMs) (The and Chin, 1988), and orthogonal Fourier Mellin moments (OFMMs) (Sheng and Shen, 1994), and generic Fourier descriptor (Zhang and Lu, 2002), wavelet moments (Shen and Ip, 1999), angular radial transforms (ART) (Bober, 2011), polar harmonic transforms (PHTs) (Yap et al., 2010), etc. Other types of region based descriptors include distribution based descriptors in which histograms are used to represent various characteristics of image. The features acquired using distribution based descriptors belong to an interest point (keypoint) or an interest region. The distribution based methods include scale invariant feature transform (SIFT) (Lowe, 2004) PCA-SIFT (Ke and Sukhtankar, 2004) gradient location and oriented histograms (GLOH) (Mikolajczyk, Schmid, 2005) histograms of oriented gradients (HOG) (Dalal and Triggs, 2005) speeded up robust features (SURF) (Bay et al., 2008), etc. The distribution based descriptors are also computation intensive and produce very high dimensionality of features.

Local Descriptors

Contour based or discrete descriptors include Fourier descriptor (Zhang and Lu, 2003) contour flexibility (Xu et al., 2009) contour point distribution histograms (Shu and Wu, 2011) Weber’s local descriptor (Chen et al., 2001) local binary patterns (LBP) (Ojala and Pietikainen, 2002) local ternary patterns (LTP) (Tan and Triggs, 2010), etc. These descriptors do not consider entire image as a whole. In fact, only a part of the shape such as boundary or image masks of various dimensions usually 978-1-5225-2848-7.ch017.m01 are considered for computing image features. These techniques provide local characteristics of the image. They are computationally efficient as compared to region based approaches.

Key Terms in this Chapter

Recall: Recall is the ability to retrieve relevant images.

Moments: Moments are scalar quantities used to characterize a function and to capture its significant features.

Precision: Precision measures the retrieval accuracy.

Complete Chapter List

Search this Book:
Reset