• No results found

Face Recognition Using Featured Histogram

N/A
N/A
Protected

Academic year: 2020

Share "Face Recognition Using Featured Histogram"

Copied!
6
0
0

Loading.... (view fulltext now)

Full text

(1)

International Journal of Emerging Technology and Advanced Engineering

Website: www.ijetae.com (ISSN 2250-2459,ISO 9001:2008 Certified Journal, Volume 3, Issue 8, August 2013)

142

Face Recognition Using Featured Histogram

Madhulakshmi

1

, Abdul Wahid Ansari

2

1Postgraduate student, G.B Pant Engineering College 2

Graduate student, UPES

Abstract Face recognition system is a computer

application for automatically identifying or verifying a person from a digital image from a source. One of the ways to do this is by comparing selected facial features from the image and a facial database. Face recognition involves three steps: Face Detection, Feature Extraction and Matching. Face detection method detect face area from the image using Viola-jones with computer vision, and some feature like eye, nose and mouth is extracted. The approach uses a set of local features that are easy to calculate and robust to partial occlusions. Histogram for each feature is calculated and used as a matching factor for recognition. The paper explores the use of local feature histograms for feature-based recognition of objects from database images. Since it only requires the calculation of very simple features, it is extremely fast and achieves more accurate recognition performance. The results show that the proposed method is with precision of 97% for the recognition of faces.

Keywords—Face detection, Feature extraction, Histogram

matching, Face recognition.

I. INTRODUCTION

The face is the feature which best distinguishes a person and face recognition is a computer application for automatically identifying or verifying a person from a digital image from a source. It is a highly active area of research with a wide variety of real-world applications. Face recognition systems are an important field in computer vision and are currently used to monitor for dangerous persons and track criminals and for other applications. A face recognition system extracts various features of face. The person identified based on these extracted features and finally the system produces the results in terms of identified and non-identified. Face Recognition has various stages such as Face detection, Feature extraction and Facial recognition.

Face detection, as a critical aspect of automatic face recognition system [2]. The purpose of face detection is to determine whether the image exist human faces. Face detection is based on identifying and locating a human face in images regardless of color, size, position, and features. There are so many Approaches for Face detection Face detection proposed by Viola and Jones [1] is most popular among the face detection approaches based on statistic methods. Various approaches have been proposed for face detection in images.

Knowledge-based methods:

These are rule-based methods. They try to capture our knowledge of faces, and translate them into a set of rules. It’s easy to guess some simple rules [3]. Knowledge-based methods extrapolate the human understanding of the structural characteristics of a face. Rules are formalized from morphological facts [12]. Facial features could be the distance between eyes or the colour intensity difference between the eye area and the lower zone. The presence of two eyes which are symmetrical, a triangular nose, the relative difference in color between parts of the human face etc. is encoded as relationships. If the rules encoded are too specific, the detection accuracy drops as small variations in the requirements results in false negatives. If the rules are too general, many false positives result. To allow for variations in the structural descriptions, a few approaches employ fuzzy theory. The approach typically builds on multi-resolution images. Initially, image analysis is carried out at a very low resolution to probe for any high level morphological description. Subsequently, the resolution is increased and the descriptions are more specific.

Feature invariant approaches:

Algorithms try to find invariant features of a face despite its angle or position. The idea is to overcome the limits of our instinctive knowledge of faces. Feature-based methods are those that take a bottom-up approach by locating facial features initially and then collecting their respective enclosing entities such as edges, blobs, streaks and graphs as a detected face [13]. Typically, edge detectors are used to identify particular shapes such as eyebrows, eyes, noses etc. and statistical models estimate distances between these shapes. A linking stage follows that collects them into groups and subsequently detects a face.

Template matching:

(2)

International Journal of Emerging Technology and Advanced Engineering

Website: www.ijetae.com (ISSN 2250-2459,ISO 9001:2008 Certified Journal, Volume 3, Issue 8, August 2013)

143 Scassellati [14] uses ratio templates for quick detection of faces. The approach virtually divides the human face into 16 regions of interest using a 14 × 16 pixels grayscale window. Each region is averaged using a grayscale window.

Appearance-based methods:

Appearance-based methods rely on techniques from statistical analysis and machine learning to find the relevant characteristics of face images. Some appearance-based methods work in a probabilistic network. An image or feature vector is a random variable with some probability of belonging to a face or not. The models are learned from a set of training images that capture the representative variability of faces [15].Appearance-based attempt to reduce the dimension of the training data before extracting statistical properties. Typically, these models calculate the class-conditional probability for a feature vector extracted from the input image. If the feature vector is x, then we calculate the Bayesian likelihoods p(x|face) and p (x|non- face) from the training data [16]. Since a direct computation of likelihoods is computationally expensive techniques such as Principal Component Analysis (PCA) are used for dimensionality reduction. Eigenfaces [17] is a pioneering appearance-based approach [18].

Feature extraction process can be defined as the procedure of extracting relevant information from a face image. This information must be valuable to the later step of identifying the subject[3].A feature extraction algorithm extracts features from the data. It creates those features based on combinations of the original data. Feature extraction involves simplifying the amount of resources required to describe a large set of data accurately. Analysis with a large number of variables generally requires a large amount of memory and computation power or a classification algorithm which over fits the training sample and generalizes poorly to new samples. Feature extraction is a general term for methods of constructing combinations of the variables to get around these problems while still describing the data with sufficient accuracy.

Face recognition is a method to recognise the face from a database. This is an evolving area, changing and improving constantly. Many research areas affect face recognition - computer vision, optics, pattern recognition, neural networks, machine learning. It has so many Different approaches in which we discuss here some of them.

Geometric/Template Based approaches:

The template based methods compare the input image with a set of templates. The set of templates can be constructed using statistical tools like Support Vector Machines (SVM), Principal Component Analysis (PCA), Linear Discriminant Analysis (LDA), Independent Component Analysis (ICA), Kernel Methods etc. [19, 20].The basis of the template matching strategy is to extract whole facial regions (matrix of pixels) and compare these with the stored images of known individuals[16]. Once again Euclidean distance can be used to find the closest match. However, there are far more sophisticated methods of template matching for face recognition. These involve extensive pre-processing and transformation of the extracted grey-level intensity values.

Piecemeal/Wholistic approaches:

Faces can often be identified from little information. Some algorithms follow this idea, processing facial features independently. The relation between the features or the relation of a feature with the whole face is not taken into account. Many early researchers followed this approach, trying to deduce the most relevant features. Some approaches tried to use the eye [23], a combination of features [20], and so on. Some Hidden Markov Model (HMM) methods also fall in this category [25].

Appearance-based approaches:

Appearance-based methods represent a face in terms of several raw intensity images. An image is considered as a high-dimensional vector. Then statistical techniques are usually used to derive a feature space from the image distribution. The sample image is compared to the training set. Linear appearance-based methods perform a linear dimension reduction [24].The face vectors are projected to the basis vectors, the projection coefficients are used as the feature representation of each face image. Examples of this approach are PCA, LDA or ICA. Non-linear appearance methods are more complicate. In fact, linear subspace analysis is an approximation of a nonlinear manifold. Kernel PCA (KPCA) is a method widely used.

Model -based methods:

(3)

International Journal of Emerging Technology and Advanced Engineering

Website: www.ijetae.com (ISSN 2250-2459,ISO 9001:2008 Certified Journal, Volume 3, Issue 8, August 2013)

144 Face detection was included as an unavoidable pre-processing step for face recognition and as an issue by itself, because it presents its own difficulties and challenges, sometimes quite different from face recognition. We have soon recognized that the amount of published information is unmanageable for a short term effort, such as required of a PFC, so in agreement with the supervisor we have stopped at a reasonable time, having reviewed most conventional face detection and face recognition approaches, leaving advanced issues, such as video face recognition or expression invariances, for the future work in the framework of a doctoral research [3].

Challenges of Face recognition: This is a critical issue in face recognition like different poses, expression, background and illumination conditions to name a few, because of which the task become difficult.

Applications of Face recognition: It is widely applied in artificial intelligence, video surveillance, identity authentication, human-machine interaction and so on.

The rest of the paper is divided into 5 sections. Next section presents the brief idea about histogram. Section 3 discusses the review of related work. Section 4 gives the idea about the proposed method uses for face recognition. The proposed result and discussion are described in section 5. Finally last section 6 concludes the paper.

II. HISTOGRAM

Histogram or Frequency Histogram is a bar graph. The horizontal axis depicts the range and scale of observations involved and vertical axis shows the number of data points in various intervals that are the frequency of observations in the intervals [25]. Histograms are popular among statisticians. Though they do not show the exact values of the data points they give a very good idea about the spread of the data and shape. The histogram is a valuable tool used to view the intensity profile of an image. It provides information about the contrast and overall intensity distribution of an image. The image histogram is simply a bar graph of the pixel intensities. The pixel intensities are plotted along the x-axis and the number of occurrences for each intensity represents the y-axis.

The histogram of a digital image with gray levels in the ranges [0, L-1] is a discrete function

h(rk) = nk ,

Where rk is the k-th gray level and nk is the number of pixels in the image having gray level rk.

It is common practice to normalize a histogram by dividing each of its values by the total number of pixels in the image, denoted by n. Thus, a normalized histogram is given by p(rk) = nk/n, for k= 0, 1, ..., L-1. Loosely speaking, p(rk) gives an estimate of the probability of occurrence of gray level rk. Note that the sum of all components of a normalized histogram is equal to 1.

III. RELATED WORK

Here discussed the some related work in which various algorithms use for face detection and recognition. All these are useful and have some benefit in different approaches, on the basis of that we will do comparison which one is good in some aspects.

The neural network with the back-propagation algorithm is used for the face detection system. The projection weights of the face and non-face blocks are computed and used as the positive and negative training vectors of the neural network, respectively. The neural network consists of a hidden layer with nine hidden nodes. Design a simple face verification method to remove these false detection blocks. The main idea of the verification scheme is to consider the distribution of the edge points from the general facial features.

One of the most significant features of human face is skin colour, for the colour image, skin colour is the relatively concentrated; stable region in the image. It’s better to distinguish human face from background regions by the skin colour. A new face detection algorithm is introduced, which combined the skin colour segmentation and the Principal Component Analysis (PCA), named SCS-PCA [5].

Skin color model is one model that it needs to use the algebra or look-up table forms to express which pixel’s color belongs to the skin color, or express the similar degree between pixel’s color and skin color.

(4)

International Journal of Emerging Technology and Advanced Engineering

Website: www.ijetae.com (ISSN 2250-2459,ISO 9001:2008 Certified Journal, Volume 3, Issue 8, August 2013)

145 SCA-PCA Algorithm process utilize skin color segmentation do the rough detection got the candidate face region, magnified the threshold value appropriate to avoid missing detection in the course of rough detect and allow a certain degree of false detection, then to regard candidate face region as the input, through the PCA method to find out the feature face space, and use eigen-face Vector to expression each sample in the face database, towards a new sample work out the space distance between it and the face database.

A powerful measure called Normalized Unmatched Points (NUP) to compare gray images and discriminate facial images. NUP works by counting the number of unmatched pixels between two images after they have been suitably pre-processed.

Huttenlocher and Rucklidhe et al [8] proposed the Hausdroff Distance (HD) and Partial Hausdorff Distance (PHD) measures to compare images. Hausdroff distance makes it suitable for face recognition because such distortions occurs frequently in facial images and are usually caused due to slight variation in pose and facial expressions.

The face recognition approach based on NUP measure is different from existing Hausdroff distance based methods as it works on gt-transformed images that are obtained from gray images rather than edges images. Thus, the approach can achieve the appearance based comparison of faces. An algorithm is also presented to efficiently compute the NUP measure.

IV. PROPOSED METHOD

In proposed method system consists of various steps: face detection, feature extraction, and recognition. There is a database of features of face images which is using for recognition. An input image is use for the detection of face from that database of image and displayed in a separated window with histogram. Then local feature like eye, nose and mouth is extracted during feature extraction, all the features and histogram is stored in a feature database. Matching processing check all feature in a sequence in one by one manner, if first feature is matched then it check the next feature, otherwise it go to check the another input images. Matching of the entire feature is the main focus of work. After the entire feature matching from data base,it moves to the recognition process. Figure 1block diagram has shown all the entire process of face recognition in.

Following are the various steps which used during the face recognition system:

4.1 Registration of face in database 4.2 Face detection

4.3 Feature extraction 4.4 Face recognition

Input image Face extracted Right

eye

Left eye

Mouth Nose

Mouth histogram Nose histogram

Left eye histogram Right eye histogram

Matching Yes

Face recognized

Feature matching database

DATABASE

Figure1. Steps of Face recognition

3.1 Face image database

A database was created with the images of the testers. The database consists of 20 different subjects. The subjects were students of GBPEC Pauri and SEC Rajasthan. The subjects were asked to face the camera and no restrictions were imposed on the expression.

3.2 Face detection

(5)

International Journal of Emerging Technology and Advanced Engineering

Website: www.ijetae.com (ISSN 2250-2459,ISO 9001:2008 Certified Journal, Volume 3, Issue 8, August 2013)

146 3.3 Feature extraction

Feature extraction is the task of reducing the high dimensional training data to a set of features to investigate characteristics of the data [10, 11].Although there are many features on human face; most of them are very useful for facial expression representation. Here features of input image are extracted such as eye, nose and mouth and calculate their histogram for recognition phase. These all extracted features of input image are matched with the feature database one by one, in a nested loop and have to decide if the person is belongs to the database or not.

3.4 Face recognition

Face recognition is the process to identify or verify people from a digital image from a database source. Recognizing is done through histogram based methods. Here apply histogram calculation for face recognition, for every extracted feature, histogram of that feature is calculated and match their histogram with feature database. First we match eye, if eye will match it will check the nose histogram matching otherwise it will check the next image and leave that image and so on. After matching of the entire feature then it match the face from database. The recognition of a face person is represented in the figure 3. The algorithm given below worked for face recognition with success rate of 97%.

V. RESULT

Here face detection method detect the face, their facial feature.Extracted feature and their respective histogram are calculated. Face detection is shown in Figure2. Each feature’s Histogram is matched with database to recognize if it exits in database or not. Histogram of the detected face is shown in Figure 3 and finally Figure 4 shows the recognition process. Matching of a feature is done in a step wise that is firstly we match the eye then nose and finally mouth. The algorithm given worked for face recognition with success rate of 97%.

While Skin color detection can detect all face images, but it also contains the neck and clothing whose color is similar to the skin. So the error rate is high. The AdaBoost method shows that some of the background is taken as to human faces incorrectly. So the result is not good either.

Figure2. Detected face

(6)

International Journal of Emerging Technology and Advanced Engineering

Website: www.ijetae.com (ISSN 2250-2459,ISO 9001:2008 Certified Journal, Volume 3, Issue 8, August 2013)

147

Figure4.Recognition of face with database

VI. CONCLUSION

In this paper, we used feature-based method to recognize the face from database. We conclude with the fact that our integrated approach to the detection, feature extraction and recognition is with high success. Our detections and recognition results, independently, are extremely competitive to the best detection and recognition results. Further result can be improves by using more than one template for matching method. These other templates can have different scales and rotations. It is also possible to improve the accuracy of the matching method by hybridizing the feature-based and template-based approaches.

REFERENCE

[1] Yao-Jiunn Chen, Yen-Chun Lin "Simple Face-detection Algorithm Based on Minimum Facial Features".

[2] P. Viola and M. Jones, ―Robust real-time object detection,‖ International Journal of Computer Vision, 57(2), 137-154, 2004. [3] Proyecto Fin de Carrera "Face Recognition Algorithms".

[4] C.C. Tsai, W.C. Cheng, J.S. Taur and C.W. Tao "Face Detection Using Eigenface and Neural Network".

[5] Liying Lang, WeiweiGu "The Face Detection Algorithm Combined Skin Color Segmentation and PCA".

[6] Adityanigam "A Noval Face Recognition Approach Using Normalised Unmatched points Measure".

[7] Yao-Jiunn Chen, Yen-Chun Lin "Simple Face-detection Algorithm Based on Minimum Facial Features".

[8] Andrew W. Senior and Ruud M. Bolle "Face Recognition And Its Applications".

[9] Felipe BallesterTomás ―COMPUTER VISION AND FACE RECOGNITION‖.

[10] W. K. Pratt, Digital Image Processing. John Wiley & Sons, Inc., 2001.

[11] A. Globerson and N. Tishby, ―Sufficient dimensionality reduction,‖ J. Mach. Learn. Res., vol. 3, pp. 1307–1331, 2003.

[12] G. Yang and T. Huang, ―Human face detection in complex background,‖ Pattern Recognition, vol. 27, no. 1, pp. 53–63, 1994. [13] S. A. Sirohey, ―Human face segmentation and identification,‖

Master’s thesis, University of Maryland, 1993.

[14] B. Scassellati, ―Eye finding via face detection for a foveated, active vision system,‖ in AAAI ’98/IAAI ’98: Proceedings of the fifteenth national/tenth conference on Artificial intelligence/Innovative applications of artificial intelligence, (Menlo Park, CA, USA), pp. 969–976, American Association for Artificial Intelligence, 1998. [15] M.-H. Yang, D. J. Kriegman, and N. Ahuja, ―Detecting faces in

images: A survey,‖ IEEE Trans. Pattern Anal. Mach. Intell., vol. 24, no. 1, pp. 34–58, 2002.

[16] M. Turk and A. Pentland, ―Eigenfaces for recognition,‖ Journal of Cognitive Neuroscience, vol. 3, no. 1, pp. 71–86, 1991.

[17] P. Belhumeur, J. Hespanha, and D. Kriegman, ―Eigenfaces vs. fisherfaces: Recognition using class specific linear projection,‖ IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 19, no. 7, pp. 711–720, 1997.

[18] M. Turk, ―A random walk through eigenspace(special issue on machine vision applications),‖ IEICE transactions on information and systems, vol. 84, no. 12, pp. 1586– 1595, 2001.

[19] L. Torres. Is there any hope for face recognition? In Proc. of the 5th

International Workshop on Image Analysis for Multimedia Interactive Services, WIAMIS, Lisboa, Portugal, 21-23 April 2004. [20] R. Gross, J. Shi, and J. Cohn. Quo vadis face recognition? - the

current state of the art in face recognition. Technical report, Robotics Institute, Carnegie Mellon University, Pittsburgh, PA, USA, June 2001.

[21] R. Brunelli and T. Poggio. Face recognition: Features versus templates. IEEE Transactions on Pattern Analysis and Machine Intelligence, 15(10):1042–1052, October 1993.

[22] A. Nefian and M. Hayes. Hidden markov models for face recognition. In Proc. of the IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP’98, volume 5, pages 2721–2724, Washington,USA, May 1998.

[23] M. Nixon. Eye spacing measurement for facial recognition. Proceedings of the Society of Photo-Optical Instrument Engineers, SPIE, 575(37):279–285, August 1985.

[24] ] B. Moghaddam. Principal manifolds and probabilistic subspaces for visual recognition. IEEE Transactions on Pattern Analysis and Machine Intelligence, 24(6):780–788, June 2002.

[25] ] Sarbjeet Singh, Meenakshi Sharma and Dr.N.SureshRao "Robust & Accurate Face Recognition using Histograms" International Conference on Emerging Trends in Computer and Image Processing (ICETCIP'2011) Bangkok Dec., 2011.

References

Related documents

As a result, Haymarket places recovery from SUDs and mental illness as a top priority for our fathers and integrates the PRF program to rebuild relationships with their children

time fraction of crest phase,

If a double door is opened at the inactive leaf, the carry bar mounted on that leaf opens the active leaf far. enough for the holding arm of the co-ordinator to ensure the

Predominance of metabolism-based resistance to photosystem II (PSII), acetolactate synthase (ALS), and p -hydroxyphenylpyruvate dioxygenase (HPPD) inhibitors in a multiple-

The health delivery organization’s community service initiatives: (a) exceed the provision of just acute medical and health care services; (b) address major health-related issues

Ethnic differences in common carotid intimamedia thickness, and the relationship to cardiovascular risk factors and peripheral arterial disease:

“PMI” is a registered trade and service mark of the Project Management Institute, Inc. ©2012 Permission is granted to PMI for PMI® Marketplace

Group 2 administered with 100% Citrus auratifolia juice and group 3 administered 50% lime juice showed marked area of cellular degeneration, vascular degeneration