A Review on Algorithmic design for Eye & Face Recognition for Various Complaint using Matlab
Volumn 4

A Review on Algorithmic design for Eye & Face Recognition for Various Complaint using Matlab

Prof. Abhijit V. Warhade *1,  Prof. Pranali K. Misal *2, Prof. Ashwini V. Kale#3

#Assistant Professor, E & C Department,PIET,RTMNU, Nagpur              


2 misal.pranali20@gmail.com

3 ashwinik734@gmail.com


It has proved that face recognition & eye scanning. It proves as strongest authentication in many electronic components as well as in electronic systems. Addhar also utilizes eye scanning which may be used as password nowadays, and face recognition system is based on feature extraction. Here we Present study regarding different algorithm available for feature extraction  & eye scanning. Different devices are used for this purpose. Here we discuss about most efficient algorithm for feature extraction with the help of MatLab. Different applications based on these algorithm are also discussed. The algorithms are Principle Component Analysis (PCA),  Linear Discriminant Analysis (LDA), skin colour, wavelet and Artificial Neural Network (ANN).

Keywords— Face Recognition system, Demanding situations in face detection and recognition.


For face recognition & eye recognition we require The input of a face recognition system is image or video stream. The output is an identification or verification of the subject or subjects that appear in the image or video.  We proposed a system which is used for various complaint using addhar card link. Facial recognition provides versatile biometric verification – face recognition could be a powerful technology capable of serving a broad spectrum of security applications, reminiscent of with police forces and customs services, however conjointly quickly extending its reach to a lot of business arenas. face recognition involves no contact and its implementation doesn’t need any extremely specialised tools, creating it ideal for distinguishing people in crowds and busy public areas .Face recognition is a crucial a part of the potential of human perception system and could be a routine task for humans, whereas building an analogous machine model of face recognition.
In the advanced technological epoch, security problems ar on the verge of risk as high rate of crimes under technical hands has been movement with the passage of your time. The vicinity protective projection is probably going to be of specific use in info reclamation applications. Computer vision offers a high stringent applications and outcomes specifically face detection and recognition. This space has continually become the researchers’ major focus in image analysis owing to its nature as human-face primary identification methodology. It is very interesting and becomes such a challenge to teach a machine to try to to this task. Face recognition also is one among the foremost troublesome issues in computer vision space. Face detection and recognition conjointly receives a large attention in medical field and analysis communities including biometric, pattern recognition and computer vision communities. If one needs to retrieve video forms beneath a vector
space model, then one can eventually ought to do a nearest neighbour search within the low dimensional area.

Face recognition technology has a very wide range of applications; as an example, the usage of the built-in camera of a telephone, tablet or computer, facial reputation software program can update passwords for tool and consumer account access. In law enforcement, the generation can resource within the identity of suspects, while in border control deployments it can streamline protection operations. another popular utility for facial recognition structures is get right of entry to manipulate at high-value web sites. in the business quarter, outlets and marketers are adopting the generation as a method to accumulate essential demographic information.


2.1 Illumination Problem Illumination drawback happens once same image with condition. thus person need to keep with fix lighting condition, fixed distance, same facial features and conjointly same read point. It will emerge extensively completely different once lighting condition is different. [8]

2.2 Create drawback Face recognition with completely different facial creates that’s referred to as pose problem. If face rotation created terribly massive changes in face appearance it cut back recognition rate. If person attempt to match same image with completely different facial create, it show the various result.

Face Detection-:

As the name suggests, it’s the detection of the face. during this part, faces square measure detected within the image. To find the face from the image there square measure four methods:

  1. Information- based methodology The rule based methodology uses the information of human to induce the data regarding the standard face. Usually, the rules capture the relationships between countenance to style the situation of t he options within the face.
  2. Model Matching methodology -In this, many customary patterns of a face square measure hold on within the info or the system to explain the face as an entire or the countenance singly. The link between associate input image and therefore the hold on patterns square measure evaluated for detection. These ways are used for each face localisation and detection.
  3. Look based mostly methodology In distinction to model matching, the models square measure learned from a group of coaching pictures that ought to capture the representative variability of the looks face. These learned models square measure then used for detection and square measure chiefly designed for face detection.
  4. Block rank patterns In this, a block rank pattern is generated by dividing 2 gradient magnitude pictures into nine(3×3) blocks and then a face is roughly detected by these 3×3 block rank patterns generated from the gradient magnitude pictures.


Detecting and recognizing faces are challenging as faces have a wide variability in poses, shapes, sizes and texture. The problems or challenges in face detection and popularity are indexed as observe [9]:

  1. Pose – A face can range depends on the placement of the digicam at some stage in the photograph is captured.
  2. Presence of structural additives-  There may be some other extra components on the face which include spectacles, moustache or beard.
  3. Those components may have different kinds, shapes, colorings and textures.
  4. Facial features- The facial features resembles at once at the individual’s face.

IV. Existing Algorithms for Face  Detection and Recognition

1. Principle Component Analysis (PCA) -:

PCA is a way in which is used to simplify the trouble of choosing the illustration of eigenvalues and corresponding eigenvectors to get a regular representation. this could be carried out through diminishing the dimension space of the illustration. to be able to gain rapid and robust item recognition, the measurement space wishes to be decreased. furthermore, PCA also retains the authentic data of the records. Eigenface based algorithm applies the PCA foundation.

2. Eigenface based algorithm -:

Eigenface based totally algorithm Eigenface based method is the most broadly used technique for face detection. in keeping with Pavan et al., eigenface is well known due to its simplicity, much less sensitive in poses and better overall performance concerning small databases or schooling sets [9]. This approach makes use of the presence of eyes, nostril and mouth on a face and relative distances among those items. This function characteristic is known as Eigenfaces in facial domain [2]. This facial characteristic may be extracted via the use of a mathematical tool known as precept component evaluation (PCA). by the usage of PCA, any original picture from the training set can be reconstructed by combining the Eigenfaces. commonly, a face is assessed as a face with the aid of calculating the relative distance of the Eigenfaces.

Fig -1.  Eigenface

V. Feature Extraction

It is the extraction of features like eyes, nose and lips from the face which can be used further to differentiate  people from each other. The approaches for face extraction are:

Fig. 2. The prominent facial features are Recognised (a) chrominance component is used. It will mouth, (b) nose, (c) eyebrow, (d) jaw line and eliminate the luminance as much as possible by (e)cheekbone
  1. DCT ( Discrete Cosine Transform )-: The Discrete Cosine remodel expresses a series of records factors in phrases of a sum of cosine functions oscillating at exclusive frequencies. therefore, it may be used to transform snap shots, compact the  versions and lets in an effective dimensionality reduction. They were broadly used for records compression.
  2. JPEG (DCT Zigzag) -: It’s far the scanning approach which actions in the zigzag f form and from low frequency thing to high  frequency factor because most of the electricity is saved in low frequency factor.
  3. Gabor Filter -: a fixed of Gabor filters with different frequencies and orientations can be useful for extracting important  functions  from an photo. They were extensively utilised in sample analysis applications.
  4. PCA (Principle Component Analysis)-:  Principle aspect evaluation is a mathematical manner that plays a dimensionality discount  by way of extracting the principle components of the multi – dimensional facts. it is based totally upon Eigenvector and a linear map.  the cases, an synthetic neural network is an adaptive gadget that modifications its shape primarily based on outside or  internal records that flows through the community.
  5. SOM (Self- Organizing Map)-: Self- Organizing Map (SOM) belongs to the aggressive studying networks. it is a kind of a neural  community that is skilled by using the usage of unsupervised studying to produce a  dimensional.

VI. Applications-:

  1. Entertainment In the area of entertainment it can be used in the monitoring of the behaviour at childcare or old people’s centres, human-robot/computer- interaction, video gaming and virtual reality etc.
  2. Information Security In the area of information Security, it can be used as personal device log on, border checkpoints, database security, file encryption, intra net security,medical records, and banks, biometric-log-in.
  3. Smart Cards  As smart cards it may be used as drivers’ licenses, passport, employee’s ID, immigration, bar code or magnetic  stripe, national ID, aadhaar Cards voter card.
  4. Law Enforcement and Surveillance  It is also used in advanced video surveillance, traffic control, ATM machines, and enhancement of CCTV  images, police bookings, suspect tracking and investigation.

VII. Conclusion-:

 This paper  work has attempted to review a considerable wide variety of  papers to cover the current improvement in face recognition  subject. present look at exposes that face reputation set of rules  may be greater the usage of hybrid techniques for better performance.  The list of references to provide more exact understanding of the procedures defined is enlisted. We apologize to  researchers whose important contributions may had been  ignored. In future scope, you could implement hybrid  method to get higher result of face reputation.


  1. CHELLAPPA, R., WILSON, C.L., and SIROHEY, S. (1995). Human and machine recognition of faces: A survey. In Proceedings of IEEE. Vol. 83, No. 5, Page. 705–740.
  2. WECHSLER, H., PHILLIPS, P., BRUCE, V., SOULIE, F., and HUANG, T. (1996). Face Recognition: From Theory to Applications. Springer-Verlag.
  3. Shou-Jen Lin, Chao-Yang Lee, Mei-hsuan Chao, Chi-Sen Chiou, Chu-Sing Yang, „The Study And Implementation Of Real-Time Face Recognition And Tracking System‟, Proceedings of the Ninth International Conference on Machine Learning and Cybernetics, Qingdao, 11-14 July 2010.
  4. Sarala A. Dabhade & Prof. Mrunal S. Bewoor „ Real Time Face Detection and Recognition using Haar – Based Cascade Classifier and Principal Component Analysis‟ International Journal of Computer Science and Management Research Vol 1 Issue 1 Aug 2012 ISSN: 2278-733X.
  5. “Blind Authentication: A Secure Crypto-Biometric Verification ProtocolManeesh Upmanyu, Anoop M. Namboodiri, Kannan Srinathan, and C. V. Jawahar”, IEEE TRANSACTIONS ON INFORMATION FORENSICS AND SECURITY, VOL. 5, NO. 2, JUNE 2010.
  6. “Very Low Resolution Face Recognition Problem” Wilman W. W. Zou, Student Member, IEEE, and Pong C. Yuen, Senior Member, IEEE, IEEE TRANSACTIONS ON IMAGE PROCESSING, VOL. 21, NO. 1, JANUARY 2012
  7. “Color Local Texture Features for Color Face Recognition” Jae Young Choi, Yong Man Ro, Senior Member, IEEE, and Konstantinos N. Plataniotis, Senior Member, IEEE , IEEE TRANSACTIONS ON IMAGE PROCESSING, VOL. 21, NO. 3, MARCH 2012.
  8. “Multibiometric Cryptosystems Based on Feature-Level Fusion” Abhishek Nagar, Student Member, IEEE, Karthik Nandakumar, Member, IEEE, and AnilK. Jain, Fellow, IEEE, IEEE TRANSACTIONS ON INFORMATION FORENSICS AND SECURITY, VOL. 7, NO. 1, FEBRUARY 2012.
  9. A Framework for Analyzing Template Security and Privacy in Biometric Authentication Systems” Koen Simoens, Julien Bringer, Hervé Chabanne, and Stefaan Seys, IEEE TRANSACTIONS ON INFORMATION FORENSICS AND SECURITY, VOL. 7, NO. 2, APRIL 2012.
  10. “A Review Paper on Biometrics: Facial Recognition” Sakshi Goel1, Akhil Kaushik2, Kirtika Goel3, International Journal of Scientific Research Engineering & Technology (IJSRET) Volume 1 Issue 5 pp 012-017 August 2012 www.ijsret.org ISSN 2278 – 0882.

Related posts

Basics of Organic and Inorganic Luminescence


Challenges & Opportunities for Automobile Vehicles in Smart Cities in India


Home Automation System


Leave a Comment