Face Detection By Open CV

ABSTRACT and the much-needed security areas in the country.

Face recognition can be used by the government to
verify the voters list, find missing persons, find the
Machine learning has been gaining momentum population or census, immigration process, also
over last decades: self-driving cars, efficient web provide security over internet scams protecting
search, speech and image recognition. The Ecommerce and highly used in the medicine and
successful results gradually propagate into our healthcare range. This brings in a very high demand
daily live. Machine learning is a class of artificial or a real time face recognition system for several
intelligence methods, which allows the computer to uses for the people and government. Providing
operate in a self-learning mode, without being such excellent systems there would be ease in
explicitly programmed. It is a very interesting and several activities. 1.2 Problem Statement The main
complex topic, which could drive the future of aim or objective of this paper is to provide or
technology. develop a system that will use the camera of the
Face detection is an important step in face computer or the system that would detect and
recognize the person’s face or the face of the
recognition and emotion recognition, which is
individual using the tool in OpenCV called as the
one of the more representative and classic Open Face and python programming language in
application in computer vision. Face is one of deep learning domain
the physiological bio-metrics based on stable
features. 2. Literature survey
Face detection by computer systems has
become a major field of interest. Face detection This section is a basic overview of the major
algorithms are used in wide range of techniques used in the face recognition system that
applications, such as security control, video apply mostly to the front face of the human being.
retrieving, biometric signal processing, human The methods include neural networks, hidden
computer interface, emotion detection, face Markov model, face matching done geometrically
recognition and image database management. and template matching. Eigenface is one of the
Face detection is a challenging mission because most widely used methods in face recognition and
faces in the images are all uncontrolled. E.g. detection which are broadly called as the principle
components in mathematical terms. The
illumination condition, vary pose, different
eigenvectors are ordered to represent different
facial expressions.
amounts of the variations in the faces. Neural
networks are highly used in the face recognition
Motivation and detection systems. An ANN (artificial neural
network) Was used in face recognition which
The most useful area in which face recognition is contained a single layer Which shows adaptiveness
important is the biometrics that is used for in crucial face recognition systems. The face
authentication process which makes the work mor verification is done using a double layer of
easier. Face recognition is one of the widely used WISARD in neural networks. Graph matching is
technologies or systems in which it has the other option for face recognition. The object as well
potential to perform tasks such as to have records as the face recognition can be formulated using
provided in by the dataset in many areas such as the graph matching performed by optimization of a
school and colleges attendance systems, it can also matching function. Hidden Markov Models is the
be helpful in catching the thieves or the terrorist, way by which stochastic modeling of nonstationary
can be helpful in the security of common people vector time series based on HMM model applied to
the human face recognition wherein the faces gets Due to the rising need for the systems which can
divided into parts such as the eyes, nose, ears, etc help in the areas such as surveillance as well as
The face recognition and correct matching is 87% security this kind of individual authentication can
correct as it always gives out the best and right no longer be done using simple handmade methods
choice of face detection through stored dataset. Or hence there is a rising need of the automated
else the relevant model reveals the identity of the systems that can easily rectify the faults and
face. The geometrical feature matching is the process the human face recognition. When the
technique which is based on the geometrical shapes work is done by machines it can perform tasks
of the face. The geometrical face configuration has efficiently in very less duration of time and cuts off
sufficient dataset for face detection and recognition the major mistakes occurred by humans. A real
system. This is one of the commonly used methods time GUI based face recognition system built can
of the face recognition and detection. This system ease this work of face detection and can be
apparently gives satisfactory results. Template achieved in various ways. 3.2 Graphical User
matching is one of the techniques through which Interface The graphical user interface (GUI) is the
the test image is represented as a two- dimensional platform that will allow the inputs from the user
array of values which can be compared using ends a kind of interaction with the system. GUI’s
Euclidean distance with single template are used in mobiles, media players, games and
representing the whole face. This method can also many others. We can design visual composition
use more than one face template from different and the temporal behaviour of the GUI in any of
points of view to represent an individual face. the software application as well as programming in
the areas of the human computer interaction. The
Methodologies GUI for this project will be widely based on the
training and the testing phase which in turn will
The concept of OpenCV was put forth by Gary allow the capture and train of the image. The
Brad ski which had the ability to perform on multi- minimum requirements for the software would be
level framework. OpenCV has a number of python along with OpenCV and the required
significant abilities as well as utilities which dataset. The minimum requirements for the
appears from the outset. The OpenCV helps in hardware would be intel i3 or any processor above
recognizing the frontal face of the person and also it and 4 core CPU. Operating systems of windows
creates XML documents for several areas such as 10 will be sufficient and random-access memory
the parts of the body. Deep learning evolved lately 8GB required. From the user end a computer or
in the process of the recognition systems. Hence laptop active internet connection and a scanner
deep learning along with the face recognition optional.
together work as the deep metric learning systems.
In short deep learning in face detection and Proposed Arrangement for system
recognition will broadly work on two areas the first design
one being accepting the solidary input image or any
other relevant picture and the second being giving In order to create this system first we will have to
the best outputs or the results of the image of the make the datasets. When the image quality
picture. We would be using dlib facial recognition becomes favourable different procedures will take
framework that would be the easy way to organize place in the face recognition system the tasks are
the face evaluation. The two main significant performed using the python queries “python
libraries used in the system are dlib and face encode_faces.py”. The input will be taken from the
recognition. Python being a very powerful dataset which will be received in the
programming languages and one of the “encodings.py”. There will be precision formatting
programming languages that are being used all in the system wherein face embedding for each face
over the world has proven to give best results in the will occur. Secondly a file
face recognition and detection systems. Together “recognize_faces_images.py” will contain all the
face recognition and detection becomes very easy required methods and the techniques for the
and fruitful with the help of the python process of identification of the face of the person
programming language and OpenCV. from the given image of the dataset. The given file
will be executed by the python command “python
Need of an automated system recognize_faces_image.py-encodings”. We can
resize or turn the image for approximity with the
goal for getting the desired output. The present
classifier along with OpenCV libraries will verify one or more persons in the scene using a
enhance the outcome or results in the face stored database of faces.
recognition system.

Project structure
Our project has four directories in the root
folder:

Dataset/: Contains our face images organized

into subfolders by name.

Images/: Contains three test images that we’ll

use to verify the operation of our model.
Figure 1: face recognition system design using
python and OpenCV. Face-detection-model/: Contains a pre-
trained Caffe deep learning model provided by
5. Advantages and Disadvantages OpenCV to detect faces. This model detects
The advantages of the face recognition system and localizes face in an image.
include faster processing, automation of the
identity, breach of privacy, massive data storage, Output/: Contains my output pickle files. If
best results, enhanced security, real time face
you’re working with your own dataset, you can
recognition of students in schools and colleges,
employees at corporate offices, smartphone unlock store your output files here as well. The output
and many more in day to day life. files include:

Few disadvantages in this system include the embedding.pickle :A Serialized facial

costing, or the funding, very good cameras of high embeddings file. Embeddings have been
definition are required, poor image quality may computed for every face in the dataset and are
limit the effectiveness of this system, size of the
image will matter because it becomes difficult to stored in this file.
recognize the face in small images, Face angles can
limit the face recognition reliability, massive le.pickle : Our label encoder. Contains the
storage is required for this system to work name labels for the people that our model can
effectively. recognize.
Problem Definition Recognizer.pickle: Our Linear Support Vector
Over the past decade face detection and recognition Machine (SVM) model. This is a machine
have transcended from esoteric to popular areas of learning model rather than a deep learning
research in computer vision and one of the better model and it is responsible for
and successful applications of image analysis and actually recognizing faces.
algorithm-based understanding. Because of the
intrinsic nature of the problem, computer vision is
not only a computer science area of research, but
also the object of neuroscientific and psychological
studies also, mainly because of the general opinion
Conclusions
that advances in computer image processing and
understanding research will provide insights into Face recognition systems are currently associated
how our brain work and vice versa. A general with many top technological companies and
statement of the face recognition problem (in industries making the work of face recognition
computer vision) can be formulated as follows: easier. The use of python programming and
given still or video images of a scene, identify or OpenCV makes it an easier and handy tool or
system which can be made by anyone according to Taiwan: IEEE,
their requirement. The proposed system discussed DOI:10.1109/ISPACS.2012.6473506, 2012.
International Symposium on Intelligent Signal
in this project will be helpful for many as it is user
Processing and Communications Systems. pp.
friendly and cost_ efficient system. Hence by the 333-337. ISBN: 978-1-4673- 5081-5.
use of python and OpenCV the face recognition
system can be designed for various purposes Face Recognition Homepage,

Acknowledgement http://www.face�rec.org/a lgorithms/

I wish to express my profound and sincere gratitude Seeing with OpenCV, Article,
to, Ms. Harshit Bhardwaj (Assistant Professor), of http://www.cognotics.com/opencv/servo
Computer Science Engineering. School of _2007_series/part_1/index.html
Computer Science Engineering, Galgotias
University, Uttar Pradesh, who guided me into the Python with deep learning,
intricacies of this project non - Chalantly with https://www.tutorialspoint.c
matchless magnanimity.
Face Detection and Tracking using OpenCV.
S.V.Viraktamath, Mukund Katti, Aditya
Khatawkar, Pavan Kulkarni. 3, s.l.: SIJ, July-
August 2013, The Standard International Journals
(The SIJ) , Vol. 1, pp. 45-50. ISSN: 2321 – 2403
References
https://www.pyimagesearch.com/2018/09/24/open
Face Detection in Real Time Based on HOG. N. J. cv-face-recognition/
Wang,S. C. Chang and P. J. Chou. Taipei,