A Project Phase-I Report on

ABANDONED OBJECT DETECTION

By

Department of Computer Science &Engineering(Data Science)

G. Narayanamma Institute of Technology & Science
(Autonomous) (For Women)
Shaikpet, Hyderabad 500 104.

Affiliated to
Jawaharlal Nehru Technological University Hyderabad
Hyderabad – 500 085
November, 2023
 A Project Phase-I Report on

ABANDONED OBJECT DETECTION

Submitted to the Department of CSE(AI & ML) and CSE(DS), GNITS in the partial
fulfillment of the academic requirement for the award of B.Tech CSE(Data Science)
under JNTUH

By

Under the Guidance of

Mrs.T.Neha
Assistant Professor, Department of CSE(AI&ML)

Department of Computer Science & Engineering (Data Science)

G. Narayanamma Institute of Technology & Science
(Autonomous) (for Women)
Shaikpet, Hyderabad- 500 104.

Affiliated to
Jawaharlal Nehru Technological University Hyderabad
Hyderabad – 500 085
November, 2023

ii
 G. Narayanamma Institute of Technology & Science
(Autonomous) (For Women)
Shaikpet, Hyderabad – 500 104.
Department of Computer Science & Engineering (Data
Science)

Certificate
This is to certify that the Project Phase-I report on “Abandoned Object Detection”
is a bonafide work carried out by Boda Swapna(20251A6732), Gajji
Meghana(20251A6742),K.ShriyaReddy(20251A6747), K.V.Satya(20251A6748) in the
partial fulfillment for the award of B.Tech degree in Computer Science &
Engineering(Data Science), G. Narayanamma Institute of Technology & Science,
Shaikpet, Hyderabad, affiliated to Jawaharlal Nehru Technological University, Hyderabad
under our guidance and supervision.

Internal Guide Head of the Department

Mrs.T.Neha Dr. O. Obulesu
Assistant Professor Associate Professor & Head
Department of CSE(AI&ML) Dept. of CSE (AI&ML) &CSE (DS)

iii
 Acknowledgements

We would like to express our sincere thanks to Dr. K. Ramesh Reddy, Principal
GNITS, for providing the working facilities in the college.

Our sincere thanks and gratitude to Dr. O. Obulesu , Associate Professor &
Head, Dept. of CSE (AI&ML) & CSE(Data Science), GNITS for all the timely support
and valuable suggestions during the period of our project.

We would like to extend our sincere gratitude to Dr. M.Seetha, Professor in CSE
and Dean-R&D, GNITS for the constant support and encouragement have motivated us
to overcome challenges during the period of our Project Phase-1.

We would also acknowledge the encouragement and support received from Dr.
N. Kalyani, Professor in CSE and Dean-Innovation and Incubation, GNITS for the
collective efforts have created a conducive learning environment and have motivated us
to push the boundaries of our capabilities.

We are extremely thankful to overall coordinator of Project Phase-1 Dr. K.

Prasanna, Assoc. Professor,Dept. of CSE(AI & ML), GNITS encouragement and
support throughout the project.

We are also extremely thankful to our project coordinators Mr. T. Rajesh, Asst.
Professor, Dept. of CSE, Mrs. Nayan Rai,Asst. Professor,Dept. of CSE(Data Science),
GNITS, for their encouragement and support throughout the project.

We are extremely thankful and indebted to our internal guide, Mrs.T.Neha,

Assistant Professor, Department of CSE(AI & ML), GNITS for her constant guidance,
encouragement and moral support throughout the project.

Finally, we would also like to thank all the faculty and staff of CSE(Data Science)
Department who helped us directly or indirectly, parents and friends for their cooperation
in completing the project work.

Boda Swapna (20251A6732)

Gajji Meghana (20251A6742)
K.Shriya Reddy (20251A6747)
K.V.Sai Satya (20251A6748)

iv
 ABSTRACT

This study presents a robust methodology for the detection of abandoned objects
within surveillance videos. The methodology leverages the precision of a YOLO-trained
pedestrian detector to identify pedestrians featured in the video sequences. By
systematically analyzing key frames captured prior to and after pedestrian interaction, the
approach intelligently discerns and locates abandoned objects, thus circumventing the
necessity for conventional background modeling techniques. Notably, the proposed
methodology boasts a general-purpose application, catering to a wide array of scenarios
beyond specific environmental settings. The experimentation conducted using the
ABODA database demonstrates the efficacy of this approach, affirming its proficiency in
identifying abandoned objects. Most strikingly, the method exhibits a heightened
resilience to fluctuations in lighting conditions, surpassing the limitations of traditional
background modeling methods. In essence, this research showcases a sophisticated and
adaptable abandoned object detection approach, poised to enhance the accuracy and
reliability of surveillance systems while accommodating real-world complexities.

Keywords:YOLO, ABODA, Background Modeling techniques, Abandoned

v
 Table of Contents
S. No Topic Page
Number Abstract..............................................................................................iv
1. Introduction....................................................................................................1
1.1. Motivation....................................................................................................2

1.2. Objectives.....................................................................................................3

1.3. Proposed Methodology................................................................................4

1.4. Hardware & Software Requirements............................................................5

2. Literature Survey..........................................................................................6
2.1. Existing System............................................................................................9

2.2. Disadvantages of Existing System...............................................................9

2.3. Problem Statement.....................................................................................10

2.4. Advantages of Proposed System................................................................10

3. Module Design...............................................................................................12
3.1. Architecture/Dataflow Diagram...................................................................12

3.2. Description of Algorithms used..................................................................14

4. Implementation of Modules..........................................................................17
4.1. Module Description......................................................................................18

4.2. Description of Technology/Tools used........................................................19

4.3. Test Case Design...........................................................................................20

5. Results & Evaluations.....................................................................................22

5.1 Description of Results....................................................................................24

6. Conclusions and Future Enhancements........................................................25

7. References.......................................................................................................26

8. Glossary............................................................................................................27
9. Appendix...........................................................................................................28

vi
 APPENDIX
YOLO - You only Look Once

OpenCV - Open Source Library for Computer

Vision CNN - Convolutional Neural Networks

RNN - Recurrent Neural Networks