Abstract social media which is totally available on

Abstract — In the digital era, most of the user’s security going to leak
due to lacking digital biometrics technology. In this paper, we are presenting
you that how to make a Face Detection and Recognition setup by using python and
open cv with High-quality Camera, this is the simple way of representing
biometrics technology. It has more advantages over other existing biometric
technologies (such as ear, fingerprint, iris, palm print etc). These parameters
include position, expression, camera quality, pose, lighting,  background, age, and gender. The main
advantage is real-time data taken without done any long process at all. In
backend we use Local Binary Pattern Algorithm, The first object comes front of
the camera then backend system capture the image and match with previous data
set by using the algorithm and predefine libraries for backend support we use
SQLite dB. We use some existing algorithms and Add with the new idea of
application belong to IOT.

Keywords—Face Detection and Recognition System, Image databases, Local Binary
Pattern, Opencv, Python, SQLite.

We Will Write a Custom Essay Specifically
For You For Only $13.90/page!


order now

 

INTRODUCTION

 

In the modern times, the face recognition software is
getting more popular in the present society of networks. In the cases like
network security , content indexing and retrival from face recognition system
make it a state of the art technology because right now every person is mostly
also have a second life on social media which is totally available on the
internet and it also made virtually impossible to the hackers to steal anyone’s
password.

Thus it also increses the human – machine interaction
process and due to simustaneous indexing and retrieving of video data it is
very usefull to the professtionals like news reporters and armed forces. For
the usage like videocalling and videoconferencing, the use of face recognition
system provides us a much efficient coding scheme. The facial recognition
system is also a

 

Biometrics software system of identifying  any individual by comparing the recent capture
or the digital image data with the stored database of images for that person to
get the output. There are many different forms of identification systems like
PIN and passwords but since  every
Biometric identifiers are totally unique to every other individual, and they
are very much consistently good in quality in  the verifying identity than any other token
and knowledge-based methods. These kind of 
systems are mostly used in the security situations. The fingerprint scan
Recognition, Facial data Recognition, Iris capture Recognition, Voice data
Recognition and Signature curve Recognition , these all are the examples of
Biometric systems. The face recognition system is a computer software which is
a system able to first identifying the person then verifying that person from a
digital image or a video snapshot from any video source. One of the ways to do
this is to by comparing diffrent selected facial features from multiple images
and a database of faces. It is mostly used in many other security systems and
it can be also compared to any other biometrics like fingerprint scanner or eye
iris scanning systems. Recently, it has also become popular in the  commercial identification and marketing. There
are different Biometric techniques, like fingerprinting and other method but the
face recognition is the most used it is most efficient.  Properly designed systems are also present at
the airports, multiplexes, and other public places where they identify the
individuals hidden in the crowd, without the nearby person even being a little
aware of the system. The other Biometrics systems which also uses the
fingerprints scanning, iris scaning, and the speech recognition system cannot
perform in this level of mass identification in such a short time. For the face
recognition system, there are two satges in which the first is verification.
This is where the system will compare the given individual with the person that
individual says they are and gives a yes or no decision by comparing them with
a database of face images. The facial recognition process normally has more of
a four stages which happen simultaneously which are processed back to back
which are face detection in the image, normalization, feature extraction, and
face recognition. The detection stage is the entry level stage which includes
identifying and the locating of the face in an image. The recognition stage is
the second stage in which feature extraction is included, where the important
information for discrimination is saved in separate area and the matching is
where the recognition result is given with the aid of a face database.

The Paradigm of the Face Recognition

Without being affected by the particular
fact that are not able to be denied  that at this
moment already varied of any business face recognition systems area unit in
use, this fashion of identification continues to be an academic degree
attention-grabbing  topic for  researchers, this is often actually because of
the particular in controvertible fact that these systems perform well to a
lower place relatively easy and controlled environments, but perform totally
worse even one factor varies in any factors area unit, like cause ,viewpoint,
facial expressions, time (when the images area unit made) and effect of  light at the time of analyze image1. The purpose
of this system is to properly analysis & process important factors and that
supports for build smart face recognition. Basic criteria used in face
recognition shown below:-

 

 
 
 
 
 
Identify
 Image

Feature
extraction

Face
Representation

                                                                                               
input image

 
 
Image Database

 
Image Classification

 

 

 

 

 

 

 

identification process(1.1)

 

The process of private identification by pattern face recognition
is going to be split into three main phases as shown in (figure one.1). These
face illustration, feature extraction and then classification2. In Face
Recognition, the first task is to identify, 
 the pattern of  face and their meta data which is uniquely
identify the input image, this is the process used to represent a face
determines the preordered algorithms used for detection and identification of
different- different categories of faces. In this level recognition, the image
is transformed  until it’s getting the
constant  ‘position’ which uniquely
identify the input image. and thanks to the data taken from image. In face
recognition, Feature extraction is the one of the important  processer, this is useful and sensitive way of
 the face image unit extracted. The input
 image is compared with another  images present in databased. This is
classification section 3, 4. The resultant is used as identity of a input face
image from the information with the absolute matching score gain, so with the
tiniest variation compared to the input face image.

 

Local Binary Patterns

 

There are number of Techniques used as extracting the most
useful data captured from photos to perform FRS. One among these options extraction
methods are that the native Binary Pattern (LBP) methodology. native binary The
initial LBP methodology is utilized to extract a texture feature that is
projected by Ojala et al. 5.It provides the native-distinction live of the
image. initial  LBP is made public in an
exceeding neighborhood of eight pixels, and gray worth center constituent.

 

Algorithm for LBP

 

Steps:-

 

1- Distribute the examined window into cells

2- Compare every single component cell to every other
one of its eight neighbours which means left-right, top-bottom and sideways. In
short compare every pixel in a circle formation.

3- If the middle pixel is bigger than than neighbor
write it as 1 otherwise 0. this provides associate degree 8-digit binary range.

4- reckon the bar graph, of the cell ,the frequency of
every “number” occurring

5-In last normalize the bar graph.

6- Concatenate the (normalized) histograms of all
cells. This also provides the feature vector for the window.

 

 

 

Formulation:-

 

 

 

So we tend to compare the middle worth with all
different neighbor block and calculate the edge perform is zero or one

 

 

 

 

                 One of the suitable model for
LBH(1.2)

 

 

0,0

1,0

2,0

 0,1

1,1

2,1

 0,2

1,2

2,2

 

 Now
we want to get LBP for pixel 1,1

 

 

157

178

220

219

218

255

215

219

255

 

We compare grey level value of pixel 1,1 with
neighbors

 

 

 

218<157==0 ,      218<178==0 ,      218<220==1 , 218<255==1 ,      218<255==1 ,      218<219==1 , 218<215==0 ,      218<219==1     Resultant the native binary pattern is for 1,1 is   0 0 1 1 1 1 0 1                                           = 61   We store and figure every LBP for texture, there are thirty-six distinctive LBPs for the image with 256 gray level, we tend to choose the variety of texture kind a category referred to as coaching information.   These choices incorporate LBP(Local Binary Pattern) it  describe the process of calculating  pixels from surrounding intervals of the regions. In LBP where it divide the overall image into (3*3) ratio from surroundings area and then  The one feature chart, that forms an illustration of the pictures that can  be compare  by measurement of the similar to keep with several studies6 FRS is exploitation the methodology provided by Local Binary Pattern which gives excellent results, every time in terms of speed  and its accuracy with performance, the variety of photos is portrayed, it's working technique looks to be quite durable against input  face photos with different angle of measuring  with different facial expressions completely different lighting conditions.   As soon as the operator is detected it is noted that it fails to encrypt the details in different scaletherefore variable neighborhood used Associate in the Nursing arbitrary variety of neighbors in a circle with a variable radius, which allows the capturing of the subsequent neighborhoods:-                                      Fig (1.3)   As shown an image, the circle is made up of R(radius) from the middle element. Where the sampling points P on the sting of this circle area unit drawn and compared with the middle element. To get the values of all P(where P is sampling points)  within the neighborhood of any radius and any variety of pixels, interpolation is important. For the neighborhoods, the notation (P, R)  used.    Figure 1.4 illustrates 3 neighbor-sets for various values of P and R. CIRCULARLY NEIGHBOR-SETS FOR 3 TOTALLY DIFFERENT VALUES OF P AND R6  fig(1.4) Software Python:- Professionally, Python is nice for backend internet development, knowledge analysis, AI, and scientific computing. several developers have additionally used Python to create productivity tools, games, and desktop apps, therefore their area unit many resources to assist you to find out how to try to those moreover. There area unit 5 most significant reasons why individuals use Python (according to language's website):- * is powerful and quick * "plays well with others" * "runs everywhere" is straightforward to find out (so you'll use it even once your journey with programming has simply begun) is friendly (thanks to the community hosts conferences and meetups, The additional helpful comes there area unit, the additional probably somebody has already engineered a performance you wish and engineered it well, which is able to greatly speed up your development method. Over 950 Python comes have over five hundred stars. Python is additionally glorious to own AN abundance of libraries that assist with knowledge analysis and scientific computing. additionally, PyGames may be a neat game engine to create games with if you would like to form easy games.   Open cv:- OpenCV (Open Source Computer Vision) is a bunch of computer vision programming functions which is used by real-time data, developed by Intel's research center supported by Willow Garage and Itseez is maintaining now. OpenCV was developed for bringing a common platform of the applications of real-time data taken for computer vision and also accelerate the utilization of economic merchandise in machine perception. OpenCV makes straightforward for businesses to change and utilize the code since it's a BSD-licensed. The library has over 2500 optimized algorithms, which incorporates a comprehensive set of each classic and progressive laptop vision and machine learning algorithms. These algorithms will be accustomed find and acknowledge faces, determine objects, classify human actions in videos, track camera movements, extract 3D models of objects, track moving objects, turn out 3D purpose clouds from stereo cameras, notice similar pictures from a picture info, sew pictures along to provide a high resolution image of a complete scene, take away red eyes from pictures taken victimization flash, follow eye movements, acknowledge scenery and establish markers to overlay it with increased reality, etc. OpenCV has over forty-seven thousand individuals in the user community and also the calculable variety of downloads prodigious fourteen million. The library is employed extensively in firms, analysis teams and by governmental bodies. It has C, C++, Python, Java & MATLAB interfaces and supports Linux, Windows, mechanical man and waterproof OS. OpenCV leans largely towards period vision applications and takes advantage of MMX and south southeast directions once offered. A full-featured CUDA and OpenCL interfaces area unit being actively developed immediately. There area unit over five hundred algorithms and concerning ten times as several functions that compose or support those algorithms. OpenCV is written mostly in C++ and encompasses a interface that works seamlessly with STL containers. Here what we are using Python 2.7 OpenCV 3.2.0 version with opencv-contrib for better performance. Visual Studio 2015 Cmake Windows 10   CONCLUSION   Facial recognition technology is now being spread all over India. This technology also has some disadvantages and weaknesses but it has a massive scope in India also. This technique is not so easy to implement and it is one of the challenging technique to be implemented. In this paper, we have given the uses of biometrics for face recognition over other traditional techniques. This paper elaborated about the use and advantages of facial recognition system and about how it actually works. Face recognition is one of the newest concept and also the most reliable one. It is user-friendly and allows the security aspect to stay in the no-risk zone. This technique is very useful in many areas of the world like crime branches, counter-terrorism, investigation field etc. The effectiveness of Facial recognition has taken the concern of people towards it. So in this paper, we have tried to give the working of facial recognition technique for security purposes.   Future Work   The future work is going on Artificial Intelligence mostly just because learning method takes time to learn and then do the further task but Artificial intelligence makes work more simple and acquire. But before proceeding to AI fully,  we will check the advantages and disadvantages of  ML and do same as AI for comparing the overall performance. FRS can be used in any place with low maintenance and I makes the very important role in security perspective there are several methods like fingerprinting and iris scanning methods if the performance of FRS is acceptable then we combine the other methods of Biometrics just for ensuring the safety of security and privacy factor with encrypt all the data extracted from the image with ANN.  In this paper, we mentation the Local binary pattern but right now there is many another method for dealing with the same situation so later we can use the Extended local binary pattern for texture classification or BOW model that is used for machine learning and classification.   References 1     W. Zhao, R. Chellappa, P. J. Phillips, and A. Rosenfeld," Face recognition: A literature survey" ACM Computing Surveys (CSUR), 35(4):399{458, 2003. 2     M. Turk and A. Pentland, "Eigenfaces for recognition", Cognitive Neuroscience, 3:72{86, 1991. 3     M. Kirby and L. Sirovich, "Application of the Karhunen-Loeve procedure for the characterization of human faces" IEEE Transactions on Pattern Analysis and Machine Intelligence, 12(1):103{108, 1990 4     S. Z. Li and A. K. Jain (eds.), "Handbook of Face Recognition" Springer-Verlag, Secaucus, NJ, 2005. 5     T. Ojala, M. Pietikainen and D. Harwood, "A comparative study of texture measures with classification based on feature distributions" Pattern Recognition vol. 29, 1996. 6     Face Recognition using Local Binary Patterns LBP. By Md. Abdur Rahim, Md. Najmul Hossain, Tanzillah Wahid & Md. Shafiul Azam 7     Mandal, Bappaditya. "Face recognition: Perspectives from the real world." In Control, Automation, Robotics and Vision (ICARCV), 2016 14th International Conference on, pp. 1-5. IEEE, 2016. 8     Phillips, P. Jonathon, Harry Wechsler, Jeffery Huang, and Patrick J. Rauss. "The FERET database and evaluation procedure for face-recognition algorithms." Image and vision computing 16, no. 5 (1998): 295-306. 9     Wagner, Andrew, John Wright, Arvind Ganesh, Zihan Zhou, Hossein Mobahi, and Yi Ma. "Toward a practical face recognition system: Robust alignment and illumination by sparse representation." IEEE Transactions on Pattern Analysis and Machine Intelligence 34, no. 2 (2012): 372-386. 10   Kherchaoui, S., and A. Houacine. "Face detection based on a model of the skin color with constraints and template matching." In Machine and Web Intelligence (ICMWI), 2010 International Conference on, pp. 469-472. IEEE, 2010. 11   Nefian, Ara V., and Monson H. Hayes. "An embedded HMM-based approach for face detection and recognition." In Acoustics, Speech, and Signal Processing, 1999. Proceedings., 1999 IEEE International Conference on, vol. 6, pp. 3553-3556. IEEE, 1999. 12   Nefian, Ara V., and Monson H. Hayes. "Face detection and recognition using hidden Markov models." In Image Processing, 1998. ICIP 98. Proceedings. 1998 International Conference on, vol. 1, pp. 141-145. IEEE, 1998. 13   Hsu, Rein-Lien, Mohamed Abdel-Mottaleb, and Anil K. Jain. "Face detection in color images." IEEE transactions on pattern analysis and machine intelligence 24, no. 5 (2002): 696-706. 14   Rowley, Henry A., Shumeet Baluja, and Takeo Kanade. "Neural network-based face detection." IEEE Transactions on pattern analysis and machine intelligence 20, no. 1 (1998): 23-38.