International Journal of Scientific Research in Engineering and Management (IJSREM)

Volume: 06 Issue: 06 | June - 2022 Impact Factor: 7.185 ISSN: 2582-3930

Real Time Object Detection with Deep Learning and OpenCV

Tulshiramji Gaikwad Patil College of Engineering and Technology, Nagpur,

Maharashtra,
Gaikwad Patil Group of Institutions, Nagpur, Maharashtra.
Corresponding Author: Prof. Jayant Adhikari

Prof. Jayant Adhikari

Dept. of Computer Science and Engineering

TGPCET Mohgaon Nagpur

KUNAL KOLHE VISHAL MANKAR

Dept. of Computer Science and Dept. of Computer Science and
Engineering Engineering
TGPCET Mohgaon, Nagpur TGPCET Mohgaon, Nagpur

KRUNAL KSHIRSAGAR ANKUSH BANSOD

Dept. of Computer Science and Dept. of Computer Science and
Engineering Engineering
TGPCET Mohgaon, Nagpur TGPCET Mohgaon, Nagpur

PRASHANT CHAUDHARI
Dept. of Computer Science and Engineering
TGPCET Mohgaon, Nagpur

ABSTRACT :

Real time object detection is an immense, vibrant and complex area of computer vision.
Assuming there is a single object to be distinguished in an image, it is known as Image Localization
and in the event that there are various objects in an image, then, at that point, it is Object Detection.
Mobile networks and binary neural networks are the most generally involved techniques for current
deep learning models to perform different tasks on embedded systems.
In this project, we develop a method to distinguish an item thinking about the deep learning pre-
prepared model MobileNet for Single Shot Multi-Box Detector (SSD).

Keywords: object detection,

Neural Network,Deep Learning,Embedded System, Computer Vision.

© 2022, IJSREM | www.ijsrem.com DOI: 10.55041/IJSREM14171 | Page 1

 International Journal of Scientific Research in Engineering and Management (IJSREM)
Volume: 06 Issue: 06 | June - 2022 Impact Factor: 7.185 ISSN: 2582-3930

1. Introduction model addressed the position of the

bounding box as well as the detection
confidences (from 0 to 1) for each identified
The trial results show that the Average object.
Precision (AP) of the algorithm to recognize
various classes as vehicle, person and chair
is 99.76%, 97.76% and 71.07%, separately. B. OpenCV (Open-Source computer
vision)
This improves the accuracy of behavior
OpenCV is a library of programming
detection at a handling speed which is
functions basically focused on real time
needed for the real-time location and the
computer vision. OpenCV is an open-source
necessities of day by day observing indoor
library which is useful for computer vision
and outside. The mix of MobileNet into the
applications like CCTV film analysis, video
SSD framework forms one of the center
analysis and image analysis.
parts of our work.

However, MobileNet with the effective SSD It is an incredible tool for image processing
framework has been a hot exploration point and performing computer vision tasks.
in recent times, to a great extent because of OpenCV is written by C++ and has in excess
managing the functional limits of running of 2,500 optimized algorithms.
strong neural nets on low-end devices like
cell phones/laptops to additionally expand At the point when we make applications for
the horde of conceivable outcomes with computer vision that we do not want to make
respect to real-time applications. it from scratch instead we can utilize this
library to begin focusing on real world
problems.

2. Literature Review OpenCV has a function to read video, which

A. MobileNet-SSD : is cv2.VideoCapture(). We can access
Our proposed model depends on the webcam by passing 0 as function parameter.
MobileNetSSD architecture. One reason
why we chose this architecture is on the
because that as shown in the paper [2], it
gives good object detection accuracy while
being quicker than different architectures,
for example, YOLO.
Especially, this is valid when attempting to
detect object in real time in low computing
devices as in our system. MobileNet-SSD
permits to lessen the detection time by
addressing the model utilizing 8-bit integers
rather than 32-bit floats. Figure 1: SSD-based detection with
MobileNet as backbone.
The input of the model was set to an image
with 300 by 300 pixels and the result of the

© 2022, IJSREM | www.ijsrem.com DOI: 10.55041/IJSREM14171 | Page 2

 International Journal of Scientific Research in Engineering and Management (IJSREM)
Volume: 06 Issue: 06 | June - 2022 Impact Factor: 7.185 ISSN: 2582-3930

3. Methodology

In the Proposed System, we are going to

detect objects in real time with the help of
Mobilenet-SSD model in fast and efficient
way. We will create the Python script for
object detection using deep neural network
with OpenCV 3.4.

First, Input will be given through Realtime

video by camera or webcam, based on
streamlined MobileNet Architecture which
uses depth-wise separable convolutions to Figure 2: Proposed System Architecture
build light weight deep neural Networks. Diagram

The input video divided into frames and pass

it to MobileNet layers.

Each feature value is determined as a

4.Future Work
difference between the amount of pixel
intensity under the bright region and the Self-Driving Cars:
pixel intensity under the dark area.
Advanced management systems interpret
Every one of the possible sizes and area of sensory info to spot acceptable navigation
the image is utilized to compute these methods, similarly as obstacles and once the
elements. image sensing element detects any sign of a
living being in its path, it mechanically
An image may contain irrelevant features
stops. This happens at a awfully quick rate
and few relevant characteristics that can be
and could be a massive step towards
used to detect the object.
Driverless Cars.
The job of the MobileNet layers is to change
over the pixels from the input image into
highlights that describe the contents of the Working Screenshots:
image.

Then it passes to MobileNet-SSD model to

determine the bounding boxes and
corresponding class(label) of objects.

After that the only last step is to show or

display the Output.

Image : 1.1

© 2022, IJSREM | www.ijsrem.com DOI: 10.55041/IJSREM14171 | Page 3

 International Journal of Scientific Research in Engineering and Management (IJSREM)
Volume: 06 Issue: 06 | June - 2022 Impact Factor: 7.185 ISSN: 2582-3930

Security : Conclusion :

Object Detection plays a awfully vital role In this project, we proposed a deep learning
in Security. It is conjointly employed by the model to identify progressively the place of
government to access the protection feed the object in pictures.
and
The framework could distinguish the item
match it with their existing information to with a normal accuracy like other best in
search out any criminals or to observe the class frameworks.
robbers’ vehicle.
In this way, we utilize an object detection
Industrial Quality Check: module that can recognize what is in the real
time video stream.
Object detection is additionally utilized in
industrial processes to spot product. To carry out the module, we join the
MobileNet and the SSD framework for a
Finding a particular object through visual quick and productive deep learning-based
review may be a basic task that's concerned strategy for object detection.
in multiple industrial processes like sorting,
inventory management, machining, quality In future work, we will keep on enhancing
management, packaging etc. our detection network model, including
lessening memory utilization and speeding
Inventory management is terribly difficult as up and additionally we will add more
things area unit arduous to trace in real time. classes.
Automatic object tally and localization
permits rising inventory accuracy.

Reference:

[1] Yundong Zhang, Haomin Peng

haomin and Pan Hu, “Towards Realtime
Detection and Camera Triggering,” CS341.

[2] Ibai Gorordo Fernandez and

Chikamune Wada, “Shoe Detection Using
SSD-MobileNet Architecture,”2020 IEEE
2nd Global Conference on Life Sciences and
Technologies (LifeTech 2020).

[3] Yu-Chen Chiu, Chi-Yi Tsai, Mind-

Image :1.2 Da Ruan, Guan-Yu Shen and TsuTian Lee,
“Mobilenet-SSDv2: An Improved Object
Detection Model for Embedded Systems,”
©2020 IEEE.

© 2022, IJSREM | www.ijsrem.com DOI: 10.55041/IJSREM14171 | Page 4

 International Journal of Scientific Research in Engineering and Management (IJSREM)
Volume: 06 Issue: 06 | June - 2022 Impact Factor: 7.185 ISSN: 2582-3930

[4] Andres Heredia and Gabriel Barros- [13] Andrew G. Howard, and Hartwig
Gavilanes,” Video processing inside Adam, “MobileNets: Efficient
embedded devices using SSD-Mobilenet to Convolutional Neural Networks for Mobile
count mobility actors,” 978- 1-7281-1614- Vision Applications”, Google Inc., 17 Apr
3/19 ©2019 IEEE. 2017.

[5] G. Bradski and, A. Kaehler,

“Learning OpenCV”, OReilly Publications,
2008.

[6] Animesh Srivastava1, Anuj Dalvi2,

Cyrus Britto3, Harshit Rai4, Kavita
Shelke5,” Explicit Content Detection using
Faster R-CNN and SSD MobileNet v2,” e-
ISSN: 2395-0056 © 2020, IRJET.

[8] R. Huang, J. Pedoeem, and C. Chen,

“YOLO-LITE: A Real-Time Object
Detection Algorithm Optimized for Non-
GPU Computers,” in Proceedings - 2018
IEEE International Conference on Big Data,
Big Data 2018.

[9] Moon, J. H. Kim and W. S. Lee,

"Advanced Lane Detecting Algorithm for
Unmanned Vehicle," Proceeding of the
ICCAS2003, pp. 1130-1133, 2003.

[10] H. Park, Y. J. Son and J. H. Kim,

"Design of Advanced Tele-operated Control
System for Unmanned Vehicle," Proceeding
of the ICCAS2005, pp. 915-919, 2005.

[11] Cosmin D. Pantilie, Silviu Bota,

Istvan Haller, Sergiu Nedevschi, "Real-time
Obstacle Detection Using Dense Stereo
Vision and Dense Optical Flow", 2010,
ISBN 978-1-4244-8230-6/101\$26.00.

[12] Ryuzo Okada, Yasuhiro Taniguchi,

Kenji Furukawa, Kazunori Onoguchi,
"Obstacle Detection Using Projective
Invariant and Vanishing Lines", IEEE
International Conference in Computer
Vision (ICCV, vol. 2, 2003.

© 2022, IJSREM | www.ijsrem.com DOI: 10.55041/IJSREM14171 | Page 5