FACE RECOGNITION SYSTEM, Thesis of Computer Science

Download FACE RECOGNITION SYSTEM and more Thesis Computer Science in PDF only on Docsity!

FACE RECOGNITION SYSTEM

A Project Submitted

In Partial Fulfilment of Requirements for the Degree of

BACHELOR OF TECHNOLOGY

In

Computer Science and Technology

ABSTRACT

Face recognition is the project of figuring out an already detected item as a acknowledged or unknown face. Often the trouble of face reputation is careworn with the trouble of face detection. Face Recognition alternatively is to determine if the "face" is a person acknowledged, or unknown, the usage of for this motive a database of faces with the intention to validate this enter face. Face detection is a laptop generation that determines the vicinity and length of human faces in virtual images. The facial capabilities are detected and any other items like trees, homes and our bodies and so forth are not noted from the virtual photograph. It may be seemed as a ̳ specific ‘case of item-elegance detection, where the project is locating the vicinity and sizes of all items in an photograph that belong to a given elegance. Face detection, may be seemed as a more ̳ general ‘case of face localization. In face localization, the project is to locate the places and sizes of a acknowledged quantity of faces (normally one). Basically, there are kinds of processes to stumble on facial component withinside the given photograph i.e. function base and photograph base technique. Feature base technique attempts to extract capabilities of the photograph and healthy it in opposition to the information of the face capabilities. While photograph base technique attempts to get high-quality healthy among education and testing images.

INTRODUCTION

The face is our number one recognition of interest in social life. It performs an important function in conveying identification and emotions. We can understand some of faces found out at some point of our lifespan and become aware of faces at a look even after year SOF separation. Computational fashions of face reputation are exciting due to the fact they are able to make a contribution now no longer only to theoretical understanding however additionally realistic applications. Computers that detect and understand faces can be carried out to a huge kind of duties including crook identification, safety system, picture and movie processing, identification verification, tagging functions and human-pc interaction. Unfortunately, growing a computational version of face detection and reputation is quite hard due to the fact faces are complex, multidimensional and significant visual stimuli.

AIM OF THE PROJECT

Our aim is to provide a portable platform for face detection and recognition which is fast and easy to build and provide results in real time.

APPROACH

To create a complete venture on Face Recognition, we’ll be running on three very awesome phases: ● Face Detection and Data Gathering ● Train the Recognizer ● Face Recognition

TECHNOLOGY USED

Hardware Used Software used

Raspberry Pi model B+ Raspbian OS Buster Pi Minicamera Module 5MP 1080p OpenCV 4 MicroSD Card Python 3 A computer/laptop to flash Raspbian OS Win32Disk Imager/Rufus

What is face detection?

Face detection refers to a technology that is capable of identifying the presence of people's faces among the digital pictures. Face detection could be a broader them than the face recognition. It simply implies that a system is ready to spot that there's a face present in a picture or a video. Face detection technology are often applied to varied fields- together with security. biometrics, law social control, amusement and private safety – to supply police work and pursuit of individuals in real time.

How does it work?

Face detection algorithms generally begin by looking for human eyes one among the simplest features to find. The rule would possibly then conceive to find eyebrows, the mouth, nose, nostrils and also the iris. Once the rule concludes that it's found a facial region, it applies extra tests to substantiate that it's, in fact, detected a face. There are many ways by that we will find a face in image however the viola-jones frameworks one the most effective thanks to find faces in period which too with high accuracy, the viola-jones framework is based on coaching a model to know what's and isn't a face. Once trained, the model extracts specific options, that are then keep during a file in order that options from new pictures is compared with the antecedently keep options at numerous stages. if the image beneath study passes through every

 Mobile phone manufacturers in products. Apple 1st used facial quality to unleash its iPhone x, and keeps with the iPhone xs. Face identity entification authenticates — it makes positive you’re you while you get entry to your telecall smartphone. Apple says the chance of a random face unlocking your telecall smartphone is prepared one in one million  Colleges in the classroom. Facial recognition software will, in essence, take roll. If you choose to chop category, your prof could know.  Social media firms on websites. Facebook uses Associate in Nursing rule to identify faces after you transfer a photograph to its platform.  Businesses at entrances and restricted areas. Some corporations have listed in security badges for biometric identification systems.  Religious teams at places of worship. Churches have used facial name to experiment their congregations to look who’s gift. It’s a superb manner to music regulars and not-so-regulars, as properly on assist tailor donation requests.  Retailers in stores. Retailers can combine surveillance cameras and facial popularity to test the faces of shoppers. One goal: figuring out suspicious characters and capability shoplifters.  Airlines at departure gates. You is maybe acquainted with having Associate in Nursing agent check your boarding by skip on the gate to board your flight. At least one airline scans your face.  Marketers and advertisers in campaigns. Marketers frequently don't forget such things as gender, age, and quality whereas targeting agencies for a product or plan. Facial quality is also accustomed outline those audiences even at some issue sort of a concert

What is raspberry pi 3?

The raspberry pi is a cheap computer that can run Linux os, but it also provides a set of gpio pins that allow us to control electronic component for physical computing. If we truly put, then raspberry pi is only a credit score card sized digital boardable to jogging a few mild weight software. It is made through raspberry foundation,

that pursuits to train humans computing. All over the world, humans use raspberry pi forums to study programming skills, construct hardware projects. It also can be used for plenty responsibilities that your pc does like games, phrase processing, spreadsheets and additionally to play HD video. The predominant cause of designing the raspberry pi board is, to inspire learning, experimentation and innovation for faculty degree students.

Technical specifications

What is OpenCV?

Computer imaginative and prescient is a technique through which we are able to recognize the snap shots and videos, how they are saved and the way we are able to manage and retrieve facts from them.it performs almost important function withinside the fields of self-using cars, robotics and lots of more. OpenCV It is an open-supply library for the laptop imaginative and prescient and photograph processing. By the use of it, one cn method photos and motion pictures to discover objects, faces or even handwriting of a human. It particularly geared toward real-time laptop imaginative and prescient and used to do all of the operations associated with photos. It helps many languages like C++, java, android and python. Application  Face detection in images and videos.  Detection of shapes like circles, rectangle etc. In an image.  Text recognition in images.  Modifying image quality.  Developing augmented reality apps.  Counting objects.  Self-driving cars or driverless cars.  Medical image analysis.  Gesture recognition  Human-computer interaction  Motion tracking.

1). SETTING UP THE ENVIRONMENT

1.1) requirements for set up of raspberry pi.

 Sd card :- a minimum of 8gb class 4 or class 10 micro Sd is required, but we recommend at least use a 16gb card instead of 8gb for the proper installation of os and software.  Display and connectivity cable :- any HDMI/DVI display must paintings nice as a show for the pi. For quality outcomes we must use a show with HDMI enter however we had a VGA display subsequently we bought a VGA to HDMI adaptor to attach raspberry pi and display. Any trendy USB keyboard and mice will simply nice whilst related to raspberry pi 3.  Power supply :- the raspberry pi three is powered with the aid of using micro USB as a result we want an amazing nice strength deliver which could deliver as a minimum 2a at 5v.

1.2) installing operating system on raspberry pi.

There are many operating systems available for raspberry pi e.g. Ubuntu mate and snappy ubuntu. We can even install windows 10 core on this device, but the recommended os that should be installed is Raspbian os , which is also known as raspberry pi os. To set up a raspberry pi working device photograph on the sd card, we are able to want every other laptop with an sd card reader to set up the os photograph. There are many software program to be had which may be used to write os photograph at the micro sd card e.g. Balena etcher, rufus, raspberry pi imager and so on.

1.5) remote access of the raspberry pi

 Sometimes we simply want to get admission to the raspberry pi with out connecting to an outside display/monitor. .possibly the pi is embedded in some thing like a robot, or we can also additionally need to view a few statistics from it from elsewhere. Maybe we really do not have a spare monitor! Or we simply don’t need to boom the burden on the raspberry pi’s electricity distribution chip. Therefore, one manner or the opposite we can constantly get admission to the raspberry remotely via way of means of putting in few things.  Vnc :- we have to enable this option from the raspberry’s configuration utility i.e. “raspi-config”. After enabling this option we can view the desktop of raspberry pi on the other systems remotely with the help of a software called vnc viewer.

 ssh :- similarly by enabling this in “raspi-config”, we can access the shell/command line interface from another computer remotely and perform various activities even after not being physically present on the site. There are many great tools available with which we can access the shell, the tool which we have use is “ bitvise ssh client”. This provides very user friendly interface and also enable us to share files among raspberry pi and other computer via sftp( secure file transfer protocol).

2). raspberry pi os

Raspberry pi operating system is also known as “raspbian operating system” as mentioned earlier in the report we can download the images of os from the official website of the raspberry pi and write it into a sd card with the help of a another computer. It may be a free software package based mostly on the “debian linux”, optimized for raspberry hardware. Raspbian os comes with over 35000 packages which might be put in with simply one command.

to setup our working environment, which requires us to install certain software and libraries.

3 ). Compiling and installing OpenCV on a raspberry

#3.1: expand file system : - after the fresh install of the raspberry os, we need to expand the file system such that the operating system could use the whole memory of the micro sd card. To expand file system we are required to follow these steps :-  Open terminal and type “ sudo raspi-config ”.  And then we select the “advanced options” menu item from the appeared window and then select ‘’ expand filesystem”.  Now we hit enter to “finish” option and then reboot the raspberry pi by typing “sudo reboot”.  After rebooting, file gadget ought to had been extended to consist of all of the to be had space, we canaffirm the equal via way of means of typing withinside the terminal , the subsequent command “df -h”.

#3.2: install dependencies : - 3.2.1) the first step is to update and upgrade any existing packages. $ sudo apt-get update && sudo apt-get upgrade 3.2.2) after updating and upgrading packages we need to install the cmake which helps us to configure the openCV build process. $ sudo apt-get install build-essential cmake pkg-config 3.2.3) now as we tend to are addressing the pictures and videos, thus we'd like to put in the some image i/o packages that enable US to load numerous image files and videos files formats from the disk.  $ sudo apt-get install libjpeg-dev libtiff5-dev libjasper- dev libpng12-dev.  $ sudo apt-get install libavcodec-dev libavformat-dev libswscale-dev libv4l-dev  $ sudo apt-get install libxvidcore-dev libx264-dev 3.2.4) the openCV library comes with a sub-module named highgui that's used to show pics to our display screen and construct primary guisand to assemble we wish gtk kit. $ sudo apt-get install libgtk2.0-dev libgtk-3-dev 3.2.5) plenty of operations internal of openCV (specifically matrix operations) may optimized similarly by putting in some more dependencies:

$ wget https://bootstrap.pypa.io/get-pip.py $ sudo python get-pip.py $ sudo python3 get-pip.py # 3.4.2 ) installing virtualenv and virtualencwrapper. A digital surroundings is a unique tool used to preserve the dependencies required through extraordinary tasks in separate locations through creating isolated, independent python environments for every of them. $ sudo pip install virtualenv virtualenvwrapper $ sudo rm -rf ~/.cache/pip Now that each virtualenv and virtualenv wrapper had been installed, we want to replace our ~/.profile document to consist of the subsequent strains at the lowest of the document.

virtualenv and virtualenvwrapper

Export workon_home=$home/.virtualenvs Export virtualenvwrapper_python=/usr/bin/python Source /usr/local/bin/virtualenvwrapper.sh After updating ~/.profile file we need to reload it to make sure the changes take affect and this can be done via this command- $ source ~/.profile # 3.4.3 ) creating the virtual environment We can use the following command to make virtual environment $ mkvirtualenv cv -p python # 3.4.4 ) enter into virtual environment Finally, we can enter into the virtual environment we created just now.

$ source ~/.profile $ workon cv once we enter into the virtual environment we can see (cv) text on the command line interface. # 3.5 ) installing numpy on the raspberry pi Numpy is a python library which is use for numerical processing. We can it via pip package manager. $ pip install numpy Installation of numpy takes a bit of time i.e. Up to 12 minutes. # 3.6 ) compile and install openCV Finally your system is prepared to assemble and deployation the openCV. Make positive that we are withinside the digital environment we created few moments ago.