International Journal of Innovative Research in Science

Engineering and Technology (IJIRSET)

(A Monthly, Peer Reviewed, Refereed, Scholarly Indexed, Open Access Journal)

Impact Factor: 8.699 Volume 14, Issue 4, April 2025

 |www.ijirset.com |A Monthly, Peer Reviewed & Refereed Journal| e-ISSN: 2319-8753| p-ISSN: 2347-6710|

Volume 14, Issue 4, April 2025

|DOI: 10.15680/IJIRSET.2025.1404414|

Detection of Exam Cheating through

Surveillance using Face Detection with AI
Dr.A.Chandra Suresh
Professor, Department of ECE, Sri Vasavi Institute of Engineering & Technology (Autonomous), Nandamuru, Pedana,
A.P, India
MD. Tasleemfathema, MD. Riyazurrahaman, S. Nutan, V.Saiganesh
Department of ECE, Sri Vasavi Institute of Engineering & Technology (Autonomous), Nandamuru, Pedana, A.P, India

ABSTRACT: Exam security is a crucial aspect of academic integrity, and traditional methods ofsupervisionare often
inadequateinpreventingcheating. Thisprojectproposes an AI- powered Exam Cheating Detection System that utilizes
Raspberry Pi Pico, ESP32- CAM, and a monitor, along with a database-driven approach for real-time face
detectionandanomalyrecognition.ThesystemcaptureslivevideofeedsfromESP32- CAM, processes them using AI
algorithms to detect suspicious behavior, and stores relevant data in a database for analysis. By integrating machine
learning models for facial recognition, posture analysis, and object detection, the system can identify unauthorized
individuals, multiple faces, and unusual activities during exams. The Raspberry Pi Pico acts as a processing unit to
handle sensor inputs and communication, while the ESP32-CAM transmits video data. A connected monitor displays
real-time alerts, enabling invigilators to intervene when necessary. This approach enhances the security and fairness of
examinations by leveraging AI-driven automation, reducing human bias, and improving surveillance efficiency.

KEYWORDS: Raspberry Pi Pico, Esp32 Cam, Monitor

I. INTRODUCTION

Academic integrity is foundational to the credibility and fairness of educational institutions. However, maintaining
strict examination protocols has always been a challenge, especially with growing class sizes, limited human resources,
and the increasing sophistication of cheating techniques. Traditional methods of invigilation, which primarily depend
on human supervision, are often inadequate in detecting subtle forms of misconduct such as covert communication,
impersonation, or the use of unauthorized electronic devices.

In response to these limitations, the integration of artificial intelligence (AI) and embedded systems presents a
promising solution. This project proposes an AI-powered Exam Cheating Detection System that leverages low-cost,
accessible hardware components—Raspberry Pi Pico and ESP32-CAM—in combination with intelligent software
capabilities. The objective is to create an automated, efficient, and accurate surveillance system capable of real-time
monitoring, detection of suspicious behaviors, and alerting of invigilators during exams.

By employing computer vision and machine learning techniques, the system enhances surveillance by reducing human
bias, minimizing manual intervention, and providing a scalable approach to secure exam environments. This project not
only strengthens exam integrity but also sets a precedent for smarter educational technologies in the future.

II. PROPOSEDSYSTEM

The proposed AI-powered Exam Cheating Detection System is designed to monitor and detect instances of cheating
during examinations through the integration of embedded hardware and intelligent software. It leverages real-time
video surveillance, AI-based behavior analysis, and database logging to create a robust and automated proctoring
solution.

IJIRSET©2025 | An ISO 9001:2008 Certified Journal | 8750

 |www.ijirset.com |A Monthly, Peer Reviewed & Refereed Journal| e-ISSN: 2319-8753| p-ISSN: 2347-6710|

Volume 14, Issue 4, April 2025

|DOI: 10.15680/IJIRSET.2025.1404414|
The system primarily consists of three main components: the ESP32-CAM module for video capture, the Raspberry
Pi Pico for control and communication, and a monitor for displaying real-time alerts. Together, these components
work in synchronization to detect and flag suspicious behaviors and activities in the exam hall.

Core Features of the Proposed System:

1. Live Video Surveillance:
o The ESP32-CAM captures continuous video footage of the examination environment.
o Positioned strategically, it covers multiple angles or individuals depending on the layout.
o Video data is streamed wirelessly for processing and analysis.
2. Artificial Intelligence-Based Detection:
o Facial Recognition: Compares detected faces with a database of registered students to prevent
impersonation.
o Multiple Face Detection: Flags scenarios where more than one face appears in the frame, indicating
possible collaboration or impersonation.
o Posture Analysis: Detects unnatural or suspicious body movements such as leaning, looking away
repeatedly, or gestures indicating communication.
o Object Detection: Identifies unauthorized objects like mobile phones, earbuds, or notes that may aid
in cheating.
3. Embedded System Processing:
o The Raspberry Pi Pico acts as the main microcontroller to manage sensor data, trigger alerts, and
facilitate communication between modules.
o It handles inputs from auxiliary sensors (if any) and supports decision-making based on AI
inferences.
4. Real-Time Alert System:
o A monitor connected to the system displays real-time video feeds with visual cues (such as red boxes,
pop-up messages, or sounds) when anomalies are detected.
o This allows invigilators to monitor multiple students simultaneously and take immediate action when
necessary.
5. Database-Driven Logging:
o All flagged events, along with timestamps, face IDs, and image snippets, are logged into a centralized
database.
o This allows for post-exam review, data analysis, and audit trails in case of disputes or further
investigations.
6. Scalability and Modularity:
o The system is modular and can be scaled to cover larger exam centers by adding more ESP32-CAM
units.
o Wireless communication between units ensures minimal setup time and hardware complexity.

Block Diagram:

IJIRSET©2025 | An ISO 9001:2008 Certified Journal | 8751

|www.ijirset.com |A Monthly, Peer Reviewed & Refereed Journal| e-ISSN: 2319-8753| p-ISSN: 2347-6710|

Volume 14, Issue 4, April 2025

|DOI: 10.15680/IJIRSET.2025.1404414|
III. EXPERIMENTAL RESULTS

Fig.1: Display of no cheating

Fig.2: Display of cheating Detected

IV. CONCLUSION

The implementation of an AI-driven exam cheating detection system using Raspberry Pi Pico, ESP32- CAM, and a
monitoring setup effectively enhances the integrity of examination processes. By integrating real-time face detection
and database-driven surveillance, this system provides a robust mechanism to identify and prevent unauthorized
activities. The use of AI ensures accurate and automated detection, minimizing human intervention while
maintaining fairness and transparency in examinations.

Furthermore, the incorporation of a centralized database allows for efficient data storage and retrieval, enabling
administrators to review incidents and take necessary actions. The ESP32-CAM captures live footage, while the

IJIRSET©2025 | An ISO 9001:2008 Certified Journal | 8752

 |www.ijirset.com |A Monthly, Peer Reviewed & Refereed Journal| e-ISSN: 2319-8753| p-ISSN: 2347-6710|

Volume 14, Issue 4, April 2025

|DOI: 10.15680/IJIRSET.2025.1404414|
Raspberry Pi Pico processes the data, ensuring a seamless and lightweight system. The real-time monitoring feature
displayed on a dedicated screen enhances invigilators' ability to oversee multiple examination venues
simultaneously.

Thissystemnotonlyaddressescurrentchallengesinacademicintegritybutalsosetsafoundationforfuture advancements in AI-

based surveillance. With further improvements, such as enhanced facial recognition
algorithmsandintegrationwithbiometricauthentication,thesystemcanbecomeacomprehensivesolution for secure and fair
examinations in educational institutions worldwide.

REFERENCES

[1] A. Gupta and A. Bhat, "Bluetooth camera based online examination system with deep learning", Proc. 6th Int.
Conf. Intell. Comput. Control Syst. (ICICCS), pp. 1477-1480, May 2022.
[2] M. M.Masud, K.Hayawi, S.S. Mathew, T.Michael and M. E. Barachi, "Smart online exam proctoring
assistforcheatingdetection"inLectureNotes inComputerScience(IncludingSubseriesLectureNotesin Artificial
Intelligence and Lecture Notes in Bioinformatics), Cham, Switzerland:Springer, vol. 13087, pp. 118-132, 2022.
[3] R. M. Al_airaji, I. A. Aljazaery, H. T. S. Alrikabi and A. H. M. Alaidi, "Automated cheating detection based on
video surveillance in the examination classes", Int. J. Interact. Mobile Technol. (iJIM), vol. 16, no. 08, pp. 124-
137, Apr. 2022.
[4] R. M. Alairaji, I. A. Aljazaery and H. S. Alrikabi, "Abnormal behavior detection of students in the
examinationHallfromsurveillancevideos"inAdvancedComputationalParadigmsandHybridIntelligent Computing,
Singapore:Springer, pp. 113-125, 2022.
[5] A. A. Malik, M. Hassan, M. Rizwan, I. Mushtaque, T. A. Lak and M. Hussain, "Impact of academic cheating and
perceived online learning effectiveness on academic performance during the COVID-19 pandemic among
Pakistani students", Frontiers Psychol., vol. 14, Mar. 2023.
[6] .L.McCabe,"Cheating among college anduniversitystudents:AnorthAmerican perspective", Int. J. Educ. Integrity,
vol. 1, no. 1, Nov. 2005.
[7] S.A.Butt,"Analysisofunfairmeanscasesincomputer-basedexaminationsystems",PacificSci.Rev.B Humanities
Social Sci., vol. 2, no. 2, pp. 75-79, Jul. 2016.
[8] M.RamzanandA.Abid,AutomaticUnusualActivitiesRecognitionUsingDeepLearninginAcademia, Jun. 2022,
[online] Available: https://www.academia.edu/download/74918847/pdf.pdf.
[9] T.S.KumarandG.Narmatha,"Videoanalysisformalpracticedetectioninclassroom examination", Proc. Int. Conf. Soft
Comput. Syst., vol. 397, pp. 135-146, 2016.
[10] Z. Li, Z. Zhu and T. Yang, "A multi-index examination cheating detection method based on neural network", Proc.
IEEE 31st Int. Conf. Tools Artif. Intell. (ICTAI), pp. 575-581, Nov. 2019.
[11] R. Bandi and S. Swamy, “Plant Disease Classification and Detection using CNN,” in 2022 IEEE 3rd Global
Conference for Advancement in Technology (GCAT), 2022.
[12] B. Karunanidhi, B. Nandhini, S. Ruth Jeba Kumari, M. Sirpikadevi, and M. Sugassini, “Plant disease detection
and classifi-cation using deep learning CNN algorithms,” in 2022 IEEE International Conference on Electronics,
Computing and Com-munication Technologies (CONECCT), 2022.
[13] Madhu et al., "Identification of Paddy Leaf Disease (Blast and Brown Spot) Detection Algorithm," 2021 2nd
International Conference on Secure Cyber Computing and Communications (ICSCCC), Jalandhar, India, 2021,
pp. 23-28, doi: 10.1109/ICSCCC51823.2021.9478164.

IJIRSET©2025 | An ISO 9001:2008 Certified Journal | 8753