ABSTRACT

The primary distinguishing feature of this project is that it aims to moderate

crimes by capturing and identifying criminal faces. Crime prevention and criminal
identification are the primary issues before the police personnel, but the availability
of police personnel is limited. With the advent of security technology, this project
aims to complement the programs and software that are used to detect, record, save
and store the identity and the facial features of the criminal. Real time footage of the
camera can be used to identify suspects on scene. In this project, an automated
facial recognition system for criminal databases is implemented using the well-
known Haar feature-based cascade classifier. This system will be able to detect and
recognize faces automatically in real time, and also map these faces and collect
information from the database, displaying the required criminal files as well. The
computer language used in this project is python, and the primary library is OpenCV.
OpenCV is a set of library functions that are aimed at real time computer vision.
Other notable features of this criminal identification system are face detection, face
identification, criminal database, graphical user interface, real time video probing and
criminal file uploading. The future scope of this project also includes incorporating
this along with other notable criminal security softwares, and to broaden its range.

(VI)
 TABLE OF CONTENTS

CHAPTER NO. TITLE PAGE No

ABSTRACT vii

LIST OF FIGURES ix

LIST OF ABBREVIATIONS x

1. INTRODUCTION 10

1.1 Outline of the project 10

1.2 Users and Stakeholders 11
1.2.1 Myself 11
1.2.2 Users 11
1.2.3 Use Cases 11
1.2.3.1 Camera Window 12
1.2.3.2 Trainer Program 12
1.2.3.3 Users 12
1.2.3.4 User Prompt 12
1.3 Literature review 12

2. AIM AND SCOPE OF THE PRESENT 14

INVESTIGATION

2.1 Problem Statement 14

2.2 Aim & Scope 14
2.3 Target Audience 14
2.4 Programming language 14
2.4.1 OpenCV 16
2.4.2 Deep Learning 17

2.5 Deep Learning vs Machine Learning 19

3. EXPERIMENTAL OR MATERIALS AND METHODS

ALGORITHMS USED 20

3.1 Functional Requirements 20

3.1.1 Inbuilt Camera 20

3.1.2 Detector 20
 3.1.3 Trainer 20
3.1.4 Recognizer 20
3.1.5 Graphics 20

3.2 Non- Functional Requirements 21

3.2.1 User friendly 21

3.2.2 Versatile 21

3.3 System Requirements 21

3.2.1 Software Requirements 21

3.2.2 Hardware Requirements 21

3.4 Importance 22

3.5 Use Case Diagram 22

3.6 Algorithm Used 23

3.6.1 Haar cascade classifier 23

3.6.2 Feature Extraction 23

3.7 Crime Identification And Detection 24

3.7.1 Flow chart 24

4. RESULTS AND DISCUSSION, PERFORMANCE

ANALYSIS 26

4.1 Results and Discussions 26

4.2 Performance Analysis 27

5. SUMMARY AND CONCLUSIONS 28

REFERENCES 29
APPENDIX 30
A.SCREENSHOTS 30
B.SOURCE CODE 37
C.PLAGIARISM REPORT 42
 LIST OF FIGURES

FIGURE NO. FIGURE NAME PAGE NO.

2.1 DEEP LEARNING 18

3.1 USE CASE DIAGRAM 22

3.2 FACE DETECTION VS RECOGNITION 23

3.3 FEATURE EXTRACTION 23

3.4 TYPES OF RECOGNITION ALGORITHM 24

3.7.1 FLOW CHART 25

5.1 GUI INTERFACE 30

5.2 UPLOADING IMAGES OF CRIMINAL 31

5.3 POP UP NOTIFICATION OF SUCCESSFUL

REGISTRATION 32

5.4 DATABASE OF CRIMINAL RECORD 33

5.5 VIEWING THE FOLDER OF THE CRIMINAL 34

5.6 FACE SAMPLES OF CRIMINALS 35

5.7 FINAL OUTPUT 36

78
 LIST OF
ABBREVIATIONS

ABBREVIATION EXPANSION

IDLE Integrated Development and Learning Environment

XML Extensible Markup Language

OPENCV Open Source Computer Vision Library

MYSQL Relational database that uses SQL query

 CHAPTER 1

INTRODUCTION

Human beings have recognition capabilities that are unparalleled within the
modern computing era. These are mainly because of the high degree of
interconnectivity, adaptive nature, learning skills and generalization capabilities of the
system. The human brain has numerous highly interconnected biological neurons
which, on some specific tasks, can outperform supercomputers. Advancements in
computing capability over the past few decades have enabled comparable
recognition capabilities from such engineered systems quite successfully. Early
recognition algorithms used simple geometric models, but recently the popularity
process has now matured into a science of sophisticated mathematical
representations and matching processes. Major advancements and initiatives have
propelled face recognition technology into the spotlight. Recognition technology are
often employed in a wide selection of applications. Computers that detect and
recognize faces can be applied to a good type of practical applications including
criminal identification etc. Face detection and recognition is employed in many places
nowadays, verifying websites hosting images and social networking sites. Crime
recognition and detection is achieved using technologies associated with computing.
Features extracted from a weapon are processed and compared with similarly
processed faces present within the database. If a weapon is recognized it's known or
the system may show an identical weapon existing in the database else it's unknown.
In closed-circuit television if an unknown face appears quite just once then it's stored
in a database for further recognition. These steps are very useful in criminal
identification. In general, crime recognition techniques are divided into two groups
supporting the face representation: they use appearance-based, which uses holistic
texture features and is applied to either whole-face or specific face image and
feature-based, which uses geometric facial features and geometric relationships
between them.

1.1 OUTLINE OF THE PROJECT

The main goal of this project is to identify the criminal and register into the
database and also detect the video survillences crime which is going on in real time.
The advanced bringing up of security technology, this application is bringing forward
to implement the programs and software that are used to detect, record, preserved
and store the identities of the person and the facial features of the criminal The
system will consist of several basic elements
▪ Python IDLE
▪ OpenCV
▪ Database connectivity with MYSQL
▪ Kaggle Data set
▪ GUI interface
▪ Web cam

1.2. USERS AND STAKEHOLDERS

This section will house the users and stakeholders. The users are going to be
using the crime detection application and also the stakeholders will develop,
maintain, and test the crime detection and recognition application.

1.2.1 MYSELF

I will be developing, maintaining, and testing the crime detection and

recognition application through its phases of development.

1.2.2 USERS

The users are anyone who has the program for the crime recognition and
detection. A user may also be a bystander who happens to surpass. This
person's body may have accidentally shown itself within the camera.
Nonetheless, he or she's going to even be detected if he or she possesses a
weapon.

1.2.3 USE CASES

These are the utilization cases for the client of the criminal identification and
detection program. The programmer has access to all or any of those cases
likewise.

1.2.3.1 Camera Window

 The program has code to open the inbuilt camera of a system. It works on any
camera, and also on specific camera apps. A resizable window opens and
therefore the user’s weapon is seen in color together with the background.
1.2.3.2 Trainer Program

The trainer program enables the user to coach his or her face by spotlight
before the camera. The python idle is employed to run this code. Half a second
is taken for the trainer program to run its course, during which period it takes
double the copies of the user's face and stores it within the database folder that
has a name of the criminal.

1.2.3.3 Users

The users are anyone who appears before the camera during execution time.

1.2.3.4 User Prompt

The user prompt is that the python is idle. It allows the user to enter into both the
modules. It has two buttons. One is for real time crime detection and the other is for
entering criminal information in a database.

1.3 LITERATURE REVIEW

In the year 2012 the face recognition “A Robust Face Recognition method
using edge based features”,was an IEEE Symposium on Computers and
Informatics. In this technique and algorithm on face feature extraction and face
recognition, the true positive means the portion of face image to be detected by the
system is the distance vector of the labeled edge between two vertices.

Paul Viola, Michael Jones, proposed a system on “Rapid Object Detection using a
Boosted Cascade of Simple Features”. This object detection was proposed by Paul Viola and
Michael Jones. Although it can be trained to detect a variety of object classes, it was motivated
primarily by the problem of face detection.

The characteristics of Viola–Jones algorithm which make it a good detection algorithm

are:
Robust – very high detection rate (true-positive rate) & very low false-positive rate
always.
Real time – For practical applications at least 2 frames per second must be processed.
Face detection only (not recognition) - The goal is to distinguish faces from non-faces
(detection is the first step in the recognition process).

The algorithm has four stages:

1. Haar Feature Selection
2. Creating an Integral Image
3. Adaboost Training
4. Cascading Classifiers

The features sought by the detection framework universally involve the sums of image
pixels within rectangular areas. As such, they bear some resemblance to Haar basis
functions, which have been used previously in the realm of image-based object
detection. However, since the features used by Viola and Jones all rely on more than
one rectangular area, they are generally more complex. The figure on the right
illustrates the four different types of features used in the framework. The value of any
given feature is the sum of the pixels within clear rectangles subtracted from the sum of
the pixels within shaded rectangles. Rectangular features of this sort are primitive
when compared to alternatives such as steerable filters. Although they are sensitive to
vertical and horizontal features, their feedback is considerably coarser.

Jyoti Nautiyal, Shivali Gahlot and Pawan Kumar Mishra, “An automated technique for
criminal face identification using biometric approach”, on Advances in Communication
and Control Systems 2013.This proposed system was using biometric to detect the
criminal .This takes a long process to identify the criminal by using his finger prints.
Fingerprints have been used in criminal investigations as a means of identification for
centuries. It is one of the most important tools of crime detection because of their
robustness and uniqueness. A fingerprint is the pattern of friction ridges and valleys on
the surface of a fingertip.
 CHAPTER 2

AIM AND SCOPE OF THE PRESENT INVESTIGATION

2.1 PROBLEM STATEMENT:

This project aims to make a criminal identification and crime detection program
which will successfully detect crime along with providing a useful user interface with a
database for entering criminal information. This is done by using python and opencv
together with deep learning concepts.

2.2 AIM AND SCOPE

The aim of the criminal identification system is to capture and aid police personnel
in identifying and effectively detecting weapons and crime. The scope of those
programs extends to a variety of individuals who appear within the camera.

2.3 TARGET AUDIENCE

The audience is someone who wants to use a crime recognition program that
occupies less space and is user friendly.

2.4 PROGRAMMING LANGUAGE

Python is an interpreted, dynamically-typed, object-oriented scripting language
with a number of built-in data types.
The Python interpreter works by loading a source file or reading a line typed at
the keyboard, parsing it into an abstract syntax tree, compiling the tree into
bytecode, and executing the bytecode.
While writing a software application, you need to concentrate on the standard of
its ASCII text file to simplify maintenance and updates. The syntax rules of
Python allow you to specific concepts without writing additional code. At the
same time, Python, unlike other programming languages, emphasizes on code
readability, and allows you to use English keywords rather than punctuations.
Hence, you'll be able to use Python to make custom applications without writing
additional code. The readable and clean code base will facilitate your to keep up
and update the software without putting time beyond regulation and energy.
 you'll be able to use Python for developing complex scientific and numeric
applications. Python is intended with features to facilitate data analysis and
visualization. you'll benefit from the information analysis features of Python to
form custom big data solutions without putting beyond regular time and energy.
At the same time, the info visualization libraries and APIs provided by Python
facilitate your to visualize and present data in an exceedingly more appealing
and effective way. Many python developers even use Python to accomplish AI
(AI) and tongue processing tasks.
You can use Python to make a prototype of the software application rapidly. Also,
you'll build the software application directly from the prototype just by refactoring
the Python code. Python even makes it easier for you to perform coding and
testing simultaneously by adopting test driven development (TDD) approach.
you'll be able to easily write the desired tests before writing code and use the
tests to assess the applying code continuously. The tests may also be used for
checking if the application meets predefined requirements supporting its ASCII
text file.
However, Python, like other programming languages, has its own shortcomings.
It lacks a number of the built-in features provided by other modern programming
languages. Hence, you've got to use Python libraries, modules, and frameworks
to accelerate custom software development. Also, several studies have shown
that Python is slower than several widely used programming languages including
Java and C++. you've got to hurry up the Python application by making changes
to the applying code or using custom runtime. But you'll always use Python to
hurry up software development and simplify software maintenance.

2.4.1 OPENCV

OpenCV may be a library of programming functions mainly aimed toward real-time

computer vision. Originally developed by Intel, it was later supported by Willow Garage
then Itseez. The library is cross-platform and free to be used under the open-source
BSD license. OpenCV (Open Source Computer Vision Library) is an open source
computer vision and machine learning software library. OpenCV was built to supply a
standard infrastructure for computer vision applications and to accelerate the
employment of machine perception within the commercial products. Being a BSD-
licensed product, OpenCV makes it easy for businesses to utilize and modify the code.
The library has over 2500 optimized algorithms, which incorporates a comprehensive