[雙語翻譯]人臉識別外文翻譯—人臉識別技術(shù)綜述（原文）

1、World Journal of Computer Application and Technology 1(2): 41-50, 2013 http://www.hrpub.org DOI: 10.13189/wjcat.2013.010204 Face Recognition Techniques: A Survey V.Vijayakumari Department of Electronics and Communicatio

2、n, Sri krishna College of Technology, Coimbatore, India *Corresponding Author: ebinviji@rediffmail.com Copyright © 2013 Horizon Research Publishing All rights reserved. Abstract Face is the index of mind. It is a

3、complex multidimensional structure and needs a good computing technique for recognition. While using automatic system for face recognition, computers are easily confused by changes in illumination, variation in poses

4、 and change in angles of faces. A numerous techniques are being used for security and authentication purposes which includes areas in detective agencies and military purpose. These surveys give the existing methods i

5、n automatic face recognition and formulate the way to still increase the performance. Keywords Face Recognition, Illumination, Authentication, Security 1. Introduction Developed in the 1960s, the first semi-automated

6、system for face recognition required the administrator to locate features ( such as eyes, ears, nose, and mouth) on the photographs before it calculated distances and ratios to a common reference point, which were th

7、en compared to reference data. In the 1970s, Goldstein, Armon, and Lesk used 21 specific subjective markers such as hair color and lip thickness to automate the recognition. The problem with both of these early solut

8、ions was that the measurements and locations were manually computed. The face recognition problem can be divided into two main stages: face verification (or authentication), and face identification (or recognition).T

9、he detection stage is the first stage; it includes identifying and locating a face in an image. The recognition stage is the second stage; it includes feature extraction, where important information for the discrimina

10、tion is saved and the matching where the recognition result is given aid of a face database. 2. Methods 2.1. Geometric Feature Based Methods The geometric feature based approaches are the earliest approaches to face r

11、ecognition and detection [1]. In these systems, the significant facial features are detected and the distances among them as well as other geometric characteristic are combined in a feature vector that is used to repr

12、esent the face. To recognize a face, first the feature vector of the test image and of the image in the database is obtained. Second, a similarity measure between these vectors, most often a minimum distance criterion

13、, is used to determine the identity of the face. As pointed out by Brunelli and Poggio, the template based approaches will outperform the early geometric feature based approaches [2]. 2.2. Template Based Methods The t

14、emplate based approaches represent the most popular technique used to recognize and detect faces [3]. Unlike the geometric feature based approaches, the template based approaches use a feature vector that represent th

15、e entire face template rather than the most significant facial features. 2.3. Correlation Based Methods Correlation based methods for face detection are based on the computation of the normalized cross correlation coe

16、fficient Cn [4, 5]. The first step in these methods is to determine the location of the significant facial features such as eyes, nose or mouth. The importance of robust facial feature detection for both detection and

17、 recognition has resulted in the development of a variety of different facial feature detection algorithms. The facial feature detection method proposed by Brunelli and Poggio uses a set of templates to detect the po

18、sition of the eyes in an image, by looking for the maximum absolute values of the normalized correlation coefficient of these templates at each point in test image [6, 2]. To cope with scale variations, a set of templ

19、ates at different scales was used. The problems associated with the scale variations can be significantly reduced by using hierarchical correlation. For face recognition, the templates corresponding to the significan

20、t facial feature of the test images are compared in turn with the corresponding templates of all of the images in the database, returning a vector of matching scores computed through normalized cross correlation. The

21、similarity scores of different features are integrated to obtain a global score that is used for recognition. Other similar method that use World Journal of Computer Application and Technology 1(2): 41-50, 2013 43 obj

22、ect recognition. Throughout, they present results demonstrating the illumination cone representation. 2.8. Support Vector Machine Approach Face recognition is a K class problem, where K is the number of known individua

23、ls; and support vector machines (SVMs) are a binary classification method [18]. By reformulating the face recognition problem and reinterpreting the output of the SVM classifier, they developed a SVM-based face

24、recognition algorithm. The face recognition problem is formulated as a problem in difference space, which models dissimilarities between two facial images. In difference space we formulate face recognition as a two c

25、lass problem. The classes are: dissimilarities between faces of the same person, and dissimilarities between faces of different people. By modifying the interpretation of the decision surface generated by SVM, we gen

26、erated a similarity metric between faces that are learned from examples of differences between faces. The SVM-based algorithm is compared with a principal component analysis (PCA) based algorithm on a difficult set o

27、f images from the FERET database. Performance was measured for both verification and identification scenarios. The identification performance for SVM is 77-78% versus 54% for PCA. For verification, the equal error ra

28、te is 7% for SVM and 13% for PCA. 2.9. Karhunen- Loeve Expansion Based Methods 2.9.1. Eigen Face Approach In this approach, face recognition problem is treated as an intrinsically two dimensional recognition problem [19

29、]. The system works by projecting face images which represents the significant variations among known faces. This significant feature is characterized as the Eigen faces. They are actually the eigenvectors. Their go

30、al is to develop a computational model of face recognition that is fact, reasonably simple and accurate in constrained environment. Eigen face approach is motivated by the information theory. 2.9.2. Recognition Using

31、Eigen Features While the classical eigenface method uses the KLT (Karhunen- Loeve Transform) coefficients of the template corresponding to the whole face image, the author Pentland et.al. introduce a face detection an

32、d recognition system that uses the KLT coefficients of the templates corresponding to the significant facial features like eyes, nose and mouth [20] . For each of the facial features, a feature space is built by sele

33、cting the most significant “eigenfeatures”, which are the eigenvectors corresponding to the largest eigen values of the features correlation matrix. The significant facial features were detected using the distance fro

34、m the feature space and selecting the closest match. The scores of similarity between the templates of the test image and the templates of the images in the training set were integrated in a cumulative score that meas

35、ures the distance between the test image and the training images [20,16]. The method was extended to the detection of features under different viewing geometries by using either a view-based Eigen space or a parametri

36、c eigenspace. 2.9.3. Subspace Recognition Approaches Principal Component Analysis , LDA, and Bayesian analysis are the three most representative subspace face recognition approaches [21]. In this paper, we show that t

37、hey can be unified under the same framework. In this approach the author first model face difference with three components: intrinsic difference, transformation difference, and noise. A unified framework is then cons

38、tructed by using this face difference model and a detailed subspace analysis on the three components. Then they explain the inherent relationship among different subspace methods and their unique contributions to the

39、 extraction of discriminating information from the face difference. Based on the framework, a unified subspace analysis method is developed using PCA, Bayes, and LDA as three steps. A 3D parameter space is constructe

40、d using the three subspace dimensions as axes. Searching through this parameter space, they achieve better recognition performance than standard subspace methods 2.9.4. Class Specific Linear Projection Approach A face

41、 recognition algorithm which is insensitive to lighting direction and facial expression is developed. They adopt the pattern recognition approach for faces in lambertian surface [22]. The method for projection is base

42、d on the Fisher’s Linear Discriminant and eigen face technique along with correlation and linear subspace. By this they concluded that fisher face method is best at extrapolating and interpolating over variation in l

43、ighting. 2.9.5. Active Pixels Based Approach With the recent advances in smart phones and their ease of availability to common man, researchers are exploring efficient algorithms for face recognition on mobile devices

44、 for entertainment applications [23]. The limited memory and processing power on mobile devices pose significant challenge to the satisfactory of popular face recognition algorithms like LBP (Local Binary Patterns), I

45、ndependent Component Analysis, PCA, Neural Networks etc. In this method, the author proposed a novel and efficient algorithm is proposed using Active Pixels which capture the essential local information of the facial

46、 image. The brody transform makes the approach more robust to rotational, translational invariance’s. The experiments were conducted on standard face recognition databases like FGNET age dataset and color FERET datas

47、et, Texas 3D Face Recognition Database (Texas 3DFRD). The results demonstrated that our approach reduced memory requirement by 80% and the computational time by 70% in comparison with LBP approach while yielding same