Fraudulent Detection System Review

You might also like

Download as docx, pdf, or txt
Download as docx, pdf, or txt
You are on page 1of 2

Fraudulent Detection System

Using OpenCV and Deep Learning


Aishwarya Janardhan, Anushka Pandey, KM Nidhi, Aman Yadav
Department of Computer Science and Engineering, JSS Academy of Technical Education Noida

Abstract- Banks play an important role in the needs to be stable and reliable in order to create
economic stability of a country. They are incentive for businesses to invest with loans as a
vulnerable to armed robbery where theft involves source of funding. The central role that the bank
use of lethal weapons to perpetrate violence organization has in today’s society leads to the
against a victim and obtain money. Banks have issue that if the organization is not well prepared
installed surveillance services for security but for the crisis that faces them it will not only
affect the organization as such but also the
they are not very effective when it comes to
surrounding stakeholders. Another reason why
immediately alerting the police and taking action
the banking sector is of interest is that when
before harm is inflicted. There is a lot of pressure looking at concepts such as stability, reliability
on law enforcement authorities to reduce the and trustworthiness, terms that often people
chances of armed robbery by monitoring places relate to the banking industry, you understand the
and engaging in patrols and stopping potential importance of good crisis management.
offenders. So, there is a need to build an effective Traditionally people, as well as private
system because crime techniques are improving businesses trust the bank to take care of their
and surveillance techniques are not advanced. savings with the belief that the money will be
safe and secure. When this notion of stability
breaks down, in the event of a crisis, the
customers of the banks face the choice of
I. Introduction continuing to trust their bank or to find other
financial institutions to handle their funds.
Implementing a well working model for robbery
Bank intrusion can have a considerable impact on
management plan will rebuild the confidence of
the people involved such as the employees,
the stakeholders and bolster the opinion that the
customers and police officials. We need a model
bank is reliable and secure, minimizing the
to assess the risk of bank robbery so that we can
perception of uncertainty that occurred during the
reduce the impact on people and prioritize the
intrusion.
investments in security measures. First of all,
there has been an increase of financial
Humans glance at an image and instantly know
transactions in favour of the traditional trading
what objects are in the image, where they are,
market. Providing good financial services is vital
and often these services are offered by the and how they interact. Similarly, we need a
banking industry. The link between banks and the system that can look at an image and identify
business world in today’s society makes the objects such as helmet, gun and mask or pattern
banks a central part of the well-being of the to identify that there is a possible threat of bank
economy as a whole and a source of funding that authorities within seconds so that they report to
the law enforcement authorities and ensure
safety, close all escape routes and catch the
intruder before any harm is caused. We can use
computer vision and deep learning to build this
model. Classical computer vision and machine
learning approaches for object detection usually
suffer from the slow response time. Modern
algorithms and architectures based on artificial
neural networks, such as the YOLO (You Only
Look Once) algorithm, solve this problem
without precision losses. In this paper we provide
a demonstration of the usage of the newest
YOLOv3 algorithm for the detection of bank
intrusion. We will train the network for 3 object
Fig1. Weapon detection using YOLO
classes (gun, mask and helmet) and demonstrate
the effectiveness of the approach in a variety of
conditions by training models by providing
thousands of images.

II. Object Detection

Object Detection is used to track the motion


trajectory of an object in a video. First, we use
the OpenCV's function, select ROI, to select an
object on a frame and track its motion using
built-in-tracker. Next, instead of using select
ROI, we use YOLO to detect an object in each
frame and track them by object centroid and size
comparison. Then we combine YOLO detection
with the OpenCV's built in tracker by detecting
the object in the first frame using YOLO and
tracking them using select ROI. The video
tracking is widely used in multiple purpose such
as: human computer interaction, security and
SurVivaLink, traffic control, medical imaging,
and so on.

You might also like