Tribhuvan University

Faculty of Humanities and Social Science

Live Object Detection &Recognition Application

A PROJECT PROPOSAL

Submitted to
Nepathya college
Tribhuvan University
Tilottama-2, Rupandehi

In partial fulfillment of the requirements for the Bachelors in Computer Application

Submitted by:
Keshav Pandey
Roll Number: 9
March, 2022
Table of Contents
Introduction:.....................................................................................................................1
Problem Statement...........................................................................................................1
Objectives.........................................................................................................................1
Research Methodology.....................................................................................................2
4.1 Requirement Identification:.......................................................................................2
4.1.2 Literature Review:...................................................................................................2
4.1.2 Requirement Analysis:........................................................................................3
4.1.2.1 Functional Requirements.................................................................................3
4.2 Feasibility Analysis:...............................................................................................3
4.2.1 Technical Feasibility:..........................................................................................3
4.2.2 Operational Feasibility:.......................................................................................3
4.2.3 Economic Feasibility:..........................................................................................3
4.3 High Level Design of System:...................................................................................4
4.3.1 System Flowchart................................................................................................4
4.3.2 Use Case Diagram...............................................................................................5
Gant Chart to show the project’s time planning:.............................................................6
Expected Outcome:..........................................................................................................7
Limitations.......................................................................................................................8
References:.......................................................................................................................8

ii
Introduction:
Live Object detection & Recognition is a mobile application. This application is
developed to detect and recognize the objects via the use of application. This technology
is crucial in AI [Artificial Intelligence] because it allows the computer to view the
environment by identifying objects in visual images.

Problem Statement

Live Object detection & Recognition can solve the complexities to determine what the
object actually is. All peoples might not know the different objects so as a helpdesk for
those peoples I have invented this application.
.

Objectives
This Object recognizing & detection application is focusing on scanning the objects via
the use of application and according to those scanned objects the application will display
the object name in the screen.
1) Helps to Recognize the objects
2) To make application more user friendly
Research Methodology

Methodology

High Level
Requirement Feasibility
Design of
Identification Analysis
System

Study of
Requirement Technical Operation Economical
Existing
Collection feasibility Feasibility Feasibility
System

4.1 Requirement Identification:

4.1.1 Study of Existing System:
To develop this application I have studied "Live Object Detection" application I obtained
some ideas to develop the application by observing that application.

4.1.1.1 Tools used

flutter is used for application UI & TensorFlow is used for generating algorithm.

4.1.2 Literature Review:

Object detection plays an important role in computer vision, automatic vehicles, industrial
automation etc. Detecting objects in real time is a challenging task. Deep learning in
object detection is better than traditional target detection. Deep learning methods include
Region proposal object detection algorithms wherein it generates region proposal
networks and then classify them. These include SPPnet, Region-based Convolutional
Neural Networks, FastRCNN, Faster-RCNN etc. Regression object detection algorithms
like SSD and YOLO generate region proposal networks and classify them at the same
time. This paper summarizes the various real time object detection approaches based on
YOLO (You Only Look Once).

2
4.1.3 Requirement Analysis:
4.1.3.1 Functional Requirements

It scans the details of the objects which are being scanned from the application
and displays the name of those scanned objects name

4.2 Feasibility Analysis:

4.2.1 Technical Feasibility:

The system is technically feasible to implement. The compatible platform exists

within our domain. The technology guarantees the reliability, accuracy and data
security. Both at the development site and at server where we will be hiring the
space for the android app, and also the database would it be possible to upgrade
the system after it is developed and implemented.

4.2.2 Operational Feasibility:

Simply it scans the particular object and displays the name of those respective
object. No major training and new skills are required.

4.2.3 Economic Feasibility:

The system is economically feasible to implement. There is no need of extra

hardware components and paid software to build the system. Cost of Software to
be acquired to build and run the product is a onetime cost.
4.3 High Level Design of System:
4.3.1 System Flowchart

Camera
Capture the
object

Detection of the Object

from Scene

Feature extraction from

the object

Object Detection TensorFlow

No?
––

Matching?

Yes?

Producing
output in
recognized Text

4.3.2 Use Case Diagram

4
Gantt Chart Of the project’s time planning:
Activities 1w 2w 3w 4w 5w 6w 7w 8w 9w 10w 11w 12w
Planning

Analysis

Design

Implementation

Testing

Documentation

Review

presentation * *

Expected Outcome:

6
The Outcome of this project is that this application will scan the objects around the
surrounding and predict those objects & display the detected objects name in the mobile
screen.
Limitations:
• Only available for android & iOS devices.

• Technical person is required to use this application

References:
[1] Sonnenberg, Timothy, et al. "Combined foreign object detection and live object
protection in wireless power transfer systems via real-time thermal camera analysis."

[2] Raj, Jeberson Retna, and Senduru Srinivasulu. "Object Detection in Live Streaming
Video Using Deep Learning Approach." IOP Conference Series: Materials Science and.

[3] Shetty,
Jyothi, and Pawan S. Jogi. "Study on different region-based object detection
models applied to live video stream and images using deep learning." International
Conference on ISMAC in Computational Vision and Bio-Engineering. Springer, Cham,
2018.