EAST WEST INSTITUTE OF TECHNOLOGY

BENGALURU-560091
(Affiliated to Visvesvaraya Technological University, Belgaum, Karnataka)

DEPARTMENT OF COMPUTER SCIENCE AND ENGINEERING

PROJECT INTERNSHIP PRESENTATION

ON

“WEAPON DETECTION”
MASTERS OF TECHNOLOGY
IN
COMPUTER SCIENCE & ENGINEERING

SUBMITTED BY: UNDER THE GUIDANCE OF :

Bhargavi G M PROF. RAJASHEKAR SA
1EW23SCS01 Dept. OF
CSE,EWIT
Contents
 Introduction
 Problem Statement
 Objectives
 Existing System
 Proposed System
 System Design
 High Level Design
 Implementation Modules
 Conclusion
 References
 Introduction
 When an individual carries a weapon (firearm or a knife) out in the open, it is a strong indicator of a potentially
dangerous situation. While some countries allow for open carry firearms, in such an event, it is still advisable to
grab the CCTV operators’ attention in order to assess the situation at hand. During recent years, an increase in the
number of incidents with the use of dangerous Automated methods for video surveillance have started to emerge
in recent years, mainly for the purpose of intelligent transportation systems (ITS).

 We have focused on the specific task of automated detection and recognition of dangerous situations applicable in
general for any CCTV system. The problem we are tackling is the automated detection of dangerous weapons—
knives and firearms, the most frequently used and deadly weapons. The appearance of such objects held in a hand
is an example of a sign of danger to which the human operator must be alerted.
 Introduction (Contd…)
 Closed circuit television systems (CCTV) are becoming more and more popular and are being deployed in many

offices, housing estates and in most public spaces. There are a million of CCTV cameras that are currently in

operation in India. This makes for an enormous load for the CCTV operators, as the number of camera views a

single operator can monitor is limited by human factors. The task of the CCTV operator is to monitor and control,

detect, observe, recognize and identify individuals and situations that are potentially harmful to other people and

property but it becomes harder to monitor when there are a lot of CCTV cameras.

 A solution to the problem of overloading the human operator is to apply automated image-understanding

algorithms, which, rather than substituting the human operator, alert them if a potentially dangerous situation is at

hand.
 Problem Statement
 To design and implement a system to detect the weapon and militant in the given image and

answer the question related to image.

Input: Image containing weapon and Militant. Process:

• The processing consists of identification of the individual component part of the weapon and Militant by using CNN

algorithm .

• After identification if any weapon and militants are found that will be detected.

Output: Display weapon and militant type when weapon and militant are detected it intimates the

admin side of user.

 Objectives

 Detection of human image and criminal identification

 Detection of segment containing the weapon in the detected image segment containing human

 Extraction of features of each segment

 Design of a neural network based classifier to classify a single type of weapon verses non-

weapon

 Computation image required to detect the weapon directly in a image as compared to detection

of weapon after the human detection.

 Existing System and Drawbacks:
• The existing systems does not classify normal and abnormal events leading the police to become
more reluctant to attend the crime scenes unless there was a visual verification, either by manned
patrols or by electronic images from the surveillance camera.
• The system is done with the image classification model using CNN with the concept of
sequential models and YOLOv3 model in the darknet framework

DRAWBACKS:
 The processing speed of the existing system and accuracy will be low .

 Running under the darknet framework which is made by C language

 Background junk detections were high.

 Problem Identification
• Nowadays protection for personal and personal property becoming very important. Video
surveillance gives a good role in real-time.
• Because of these needs deployment of cameras take place at every corner, video surveillance
system understand the scene and it automatically detects abnormal activities.
• To recognize the occurrence of uncommon events such as unknown or with weapon or grenades
or tankers detects in the low-resolution video simply by using statistical property, standard
deviation of moving objects.

Problem Statement
Detection and getting details about the detection is logically hard to get implement with efficient output.There
are lots of techniques to detect the different types of anomaly detection in surveillance. Conventional methods
of detecting intrusion in naked eye observation in live time detection.
 Proposed System
Here we proposes a system for militant and military object detection using YOLOv5 and Python.
The project involves collecting a large dataset of annotated images of military objects and militants,
splitting the dataset into training, validation, and testing sets, and converting the annotated images
into YOLOv5 format. The YOLOv5 model is then trained on the annotated dataset and evaluated
using various metrics such as precision, recall, and map. Once the model is trained and evaluated, it
is deployed to detect militants and military objects in real-time images or videos using Python
libraries like OpenCV or PyTorch. The proposed system can also be integrated with additional
features such as object tracking, alert notifications, or automatic response mechanisms to enhance
its functionality. The project demonstrates the potential of deep learning-based object detection for
military and defense applications and serves as a basis for further research in this area.
System Design
The system architecture for the
proposed system. The input image is
preprocessed and converted to gray
scale image to get the clear vision of
the image. Then it will be converted
into binary values. In the next step
identifies the part which needs to
proceed further. Then required feature
are extracted by In the CNN
convolution layer. By passing those
features into different layer of CNN
we get compressed image, that feature
is used for detection of weapon and
militant using SoftMax activation
function.
Convolutional Neural Network(YOLO)
 Data Flow Diagram
 A data flow diagram (DFD) is graphic representation of the
"flow" of data through an information system.
 A data flow diagram can also be used for the visualization of
data processing (structured design).
 It is common practice for a designer to draw a context level
DFD first which shows the interaction between the system
and outside entities.
 This level of preprocessing shows that the image is given as input.
 As we giving the color image so that RGB image is converted into gray scale
values to reduce complexity in the image.
 For efficient feature extraction gray scale values are converted into binary
values.
 Then the image with reduced complexity is send to the next process.
 The figure of identification shows that the image with
reduced complexity is considered as input.
 Here the region with the value of one is considered as
black that region is considered for next process.
class
diagra
m

• Class diagram describes the attributes and operations of a class and alsotheconstraints imposed on the
system. The class diagrams are widely used in the modelingofobject oriented systems because they are
the only UML diagrams, which can be mappeddirectly with object-oriented languages.
• The purpose of fig is to model the static view of an application by using class diagramare the only
diagram which can be directly mapped with object–oriented languages andthus widely used at the time
of construction.
 Modules
1.Image Pre Processing.
2.Identification
3.Feature Extraction
4.Weapon and militant Recognition
5.Militant Detection
6.Intimation
 Image Preprocessing
Image processing is a mechanism that focuses on
the manipulation of images in different ways in
order to enhance the image quality. Images are
taken as the input and output for image
processing techniques. It is the analysis of image-
to-image transformation which is used for the
enhancement of image
 Identification
In this stage identify the region which needs to proceed for further process,
it is involved in the identification of the particular region of the image that
is used for the further process like feature extraction and classification of
the images. The output of the pre-processing step is given as the input for
the identification process. This process is based on the binary values
obtained in the pre-processing step. The region with black is considered a
region of interest. The region of interest obtained by the pre-processing of
the images. That region is considered as proceeding part of the image from
which weapon and militant will be identified. The identified weapon and
militant images are given to the feature extraction process
 Feature Extraction
In this stage extract the required feature from the
identified region which is obtained from the previous
step. That region is compressed by converting a
reduced size matrix to control over fitting. The
reduction of the matrix size helps in reducing the
memory size of the images. Then the flattening process
is applied to the reduced matrix, in which the reduced
matrix is converted to a one-dimension array, which is
used for final detection.
 Weapon and Militant Recognition
 In this stage one-dimension array is used for the final classification process. The output image
obtained from feature extraction is given as input to this process. Where continuous
classification of all the features obtained from the previous stage. Artificial neural networks are
applied in this process.
 Each node of the input layer has a value from a onedimension array which represents the feature
from the extracted region.
 That is sent to the hidden layer. Multiple features are getting from the input layer and undergo
multiple iteration in the hidden layer.
 Finally get the predictive values by applying SoftMax activation function to it. Finally, get some
output values from this process and these values undergo further process.
 The highest value in the predictive value is considered as output identified as weapon and
militant. By using these methods, the weapon and militant will be detected by considering
highest accuracy values.
 Intimation
 Telebot is a bot framework for Telegram Bot API. This package provides the best of its kind API

for command routing, inline query requests and keyboards, as well as callbacks.

 Bots have legitimate uses on the popular chat apps, but are being exploited by

cybercriminals. Cybercriminals are using bots deployed in popular messaging apps Discord and

Telegram to share alerts, new research has revealed.

 Conclusion
The proposed project aimed to develop a deep learning model that can detect militants in
real-time using the YOLO algorithm. The project was implemented in five modules,
including dataset preparation, YOLOv3 model training, real-time detection, performance
evaluation, and voice announcement. The project has potential applications in the security
and defense sectors, where the detection of militants is of critical importance. The model
was trained and evaluated using precision, recall, and F1 score metrics, and the results
showed high accuracy in detecting militants in real-time. The voice announcement system
was integrated with the real-time detection module, which provided an additional layer of
security and alerted the security personnel immediately. The project has the potential to
enhance the security and defense sectors by providing a fast and accurate militant
detection system.
 REFERENCES
 Harsha Jain et.al. “Weapon and militant detection using artificial Intelligence and deep learning
for security applications” ICESC 2020.
 Arif Warsi et.al “Automatic handgun and knife detection algorithms” IEEE Conference 2019.
 Neelam Dwivedi et.al. “Weapon and militant classification using deep Convolutional neural
networks” IEEE Conference CICT 2020.
 Kumar Verma et.al. “Handheld Gun detection using Faster RCNN Deep Learning” IEEE
Conference 2019.
 https://www.techtarget.com/searchenterpriseai/definition/convolutional-neuralnetwork
 https://www.researchgate.net/
 https://towardsdatascience.com
Thank You