You are on page 1of 9

FACE MASK DETECTION

Minor Project Synopsis


(BCA 5005)

Submitted in partial fulfillment of requirement for the award of the degree of

Bachelor of Computer Applications

Submitted by
Group No : 9
1. Vipin Yadav (Univ. Roll No: 21015000252)
2. Vaibhav Yadav (Univ. Roll No: 21015000243)
3.Vaibhav Tiwari (Univ. Roll No: 21015000242)

Section: D

Under the supervision of


Mr. Akhilesh Singh

Chhatrapati Shahu Ji Maharaj University (CSJMU)


UP State University | Formerly Kanpur University
Kanpur (UP)
(2023 – 24)
1.Introduction
In the era of global health challenges, the "Real-time Face Mask Detection System"
serves as a vital tool for enforcing safety protocols and preventing the spread of infectious
diseases. This college mini project harnesses the power of Python programming language
and computer vision techniques to create an intelligent system capable of identifying
whether individuals are wearing face masks accurately.
The significance of face masks as a preventive measure against the transmission of
respiratory diseases has been well-established. However, effective implementation and
monitoring of mask-wearing practices pose challenges, particularly in public spaces. The
proposed project aims to bridge this gap by creating an intelligent, real-time system
capable of accurately detecting individuals wearing or not wearing face masks in diverse
scenarios.
Through this endeavor, students will delve into the intricacies of image processing,
machine learning, and real-time application development, gaining valuable insights into
the practical applications of Python in addressing contemporary challenges. As the project
unfolds, it will not only showcase the technical prowess of the students but also contribute
to the collective efforts aimed at fostering a safer and healthier environment. The Real-
time Face Mask Detection System represents an embodiment of technological innovation
with a profound impact on public health and safety.

2. About the Project


In the wake of global health crises, the "Real-time Face Mask Detection System" emerges
as a cutting-edge project aimed at leveraging the power of Python programming language
and computer vision technologies to enhance public safety. The project addresses the
crucial need for enforcing face mask protocols, a key preventive measure against the
spread of infectious diseases, in various public and private settings.

The system employs advanced image processing and deep learning techniques to achieve
real-time detection of individuals wearing or not wearing face masks. Through the
integration of OpenCV for video stream capture, face detection models for identifying
facial features, and a custom-trained Convolutional Neural Network (CNN) for mask
classification, the project delivers a comprehensive solution for automated mask
compliance monitoring.

➢ 2.1 Methodology & Modules

METHODOLOGY:

1.Problem Definition:

• Clearly define the problem: the need for a real-time face mask detection system to
enhance public safety and enforce health protocols.

2.Literature Review:

• Conduct a thorough literature review on existing face detection and mask detection
techniques.
• Explore relevant studies, papers, and projects related to computer vision, deep
learning, and mask detection.

3.Data Collection:

• Acquire a diverse dataset containing images of individuals with and without


masks.
• Ensure the dataset covers various scenarios, including different mask styles,
angles, and lighting conditions.

4.Data Preprocessing:

• Implement preprocessing techniques to enhance the quality of the input images.


• Normalize, resize, and augment the dataset to improve the model's robustness.

5.Face Detection:

• Utilize pre-trained face detection models (e.g., Haarcascades or MTCNN) to


identify faces in the input frames.
• Extract and isolate the regions of interest (ROIs) containing faces for further
processing.

6.Mask Classification Model:

• Design a Convolutional Neural Network (CNN) architecture for mask


classification.
• Split the dataset into training and testing sets.
• Train the model on the training set using a suitable deep learning framework (e.g.,
TensorFlow, PyTorch).
• Fine-tune the model to achieve optimal performance.

7.Real-time Integration:

• Develop a real-time application that integrates face detection and mask


classification modules.
• Utilize OpenCV for capturing video streams and processing frames in real-time.
• Implement logic to apply the trained mask classification model to identify
individuals with or without masks.

8.Performance Evaluation:

• Assess the model's performance using metrics such as accuracy, precision, recall,
and F1 score.
• Validate the system with diverse test scenarios to ensure robustness and
generalization.

9.Testing and Validation:

• Conduct thorough testing of the system in various environments and scenarios.


• Validate the accuracy of the face mask detection system against ground truth data.
➢ By following this methodology, the Real-time Face Mask Detection System aims
to provide an effective and reliable solution to the identified problem, leveraging
the capabilities of Python, computer vision, and deep learning.

MODULES:

1.Data Collection and Preprocessing Module:

• Submodule 1: Image Acquisition


➢ Functionality: Capture real-time video streams from a webcam or utilize existing
video files.
• Submodule 2: Image Preprocessing
➢ Functionality: Implement techniques to enhance the quality of input frames, such
as resizing, normalization, and noise reduction.

2.Face Detection Module:

• Submodule 1: Haarcascades or MTCNN Implementation


➢ Functionality: Use pre-trained models to detect faces in frames.
• Dependencies: OpenCV library.

3.Mask Classification Module:

• Submodule 1: Dataset Preparation


➢ Functionality: Collect and prepare a diverse dataset containing images of
individuals with and without masks.
• Submodule 2: CNN Model Training
➢ Functionality: Train a Convolutional Neural Network using a deep learning library
(e.g., TensorFlow or PyTorch).
• Dependencies: TensorFlow or PyTorch, suitable dataset.

4.Real-time Integration Module:

• Submodule 1: Face Mask Detection Integration


➢ Functionality: Combine the face detection and mask classification modules to
identify individuals wearing or not wearing masks in real-time.

5.Performance Evaluation Module:

• Submodule 1: Metrics Calculation


➢ Functionality: Evaluate the accuracy, precision, recall, and F1 score of the face
mask detection system.
• Submodule 2: Results Visualization
➢ Functionality: Display evaluation results in a clear and understandable format.

6.Future Enhancements Module:


• Submodule 1: Edge Deployment
➢ Functionality: Explore deployment options on edge devices for decentralized
monitoring.
• Submodule 2: Additional Features Integration
➢ Functionality: Investigate the integration of features like social distancing detection
or mask usage analytics.
➢ These modules provide a structured approach to developing the Face Mask
Detection project, making it easier for team collaboration and implementation.
Each submodule focuses on specific functionalities, ensuring modularity and code
reusability.

➢ 2.2 Hardware and Software Requirements

Hardware requirements:

1.Computer System:

• A modern computer system with a multi-core processor (e.g., Intel Core i5 or


equivalent) for efficient real-time processing.
• Minimum of 8 GB RAM to handle image processing and deep learning tasks.

2.Webcam:

• A high-resolution webcam capable of capturing clear and detailed video streams.


• Alternatively, an external camera can be used for better image quality.

3.Graphics Processing Unit (GPU) (Optional but Recommended):

• A dedicated GPU (NVIDIA GeForce or AMD Radeon) is recommended for


accelerated deep learning model training and real-time inference.
• CUDA-compatible GPU for TensorFlow or a compatible GPU for PyTorch can
significantly improve performance.

4.Storage:

• Sufficient storage space for the Python environment, deep learning libraries, and
the dataset.
• An SSD (Solid State Drive) is preferable for faster data read and write operations.

Software requirements:

1.Python:

• Version 3.x or higher


2.Integrated Development Environment (IDE):

• Choose a preferred IDE for Python development. Suggested options


include:
➢ PyCharm
➢ Jupyter Notebooks
➢ Visual Studio Code

3.OpenCV:

• Install the OpenCV library for computer vision tasks:

4.Deep Learning Framework:

• Select and install a deep learning framework for model development.


Recommended options include:
➢ TensorFlow

5.Dataset:

• Obtain a suitable dataset for training the face mask detection model.
Recommended datasets include:
➢ MaskedFace-Net
➢ Labeled Faces in the Wild (LFW)
➢ Kaggle Datasets

3. Advantages & Applications

➢ Advantages:

1.Public Health Enhancement:

• The face mask detection system contributes to public health by encouraging and
enforcing the use of face masks in crowded places, reducing the risk of disease
transmission.

2.Automation and Efficiency:

• Automation of mask detection through computer vision reduces the need for
manual monitoring, ensuring a more efficient and reliable process, especially in
high-traffic areas.
3.Real-time Monitoring:

• The system operates in real-time, providing immediate feedback and enabling


timely interventions when individuals are detected without masks.

4.Adaptability:

• The project can be easily adapted to different environments, such as airports,


hospitals, public transport, and educational institutions, making it a versatile
solution for diverse applications.

➢ Applications:

1.Public Spaces and Transportation Hubs:

• Deploy the face mask detection system in airports, train stations, and bus terminals
to ensure compliance with health protocols in crowded transportation hubs.

2.Educational Institutions:

• Implement the system in schools and universities to monitor mask compliance


among students, teachers, and staff, thereby promoting a safe learning
environment.

3.Healthcare Facilities:

• Use the system in hospitals and clinics to ensure that both healthcare professionals
and patients adhere to safety measures, minimizing the risk of cross-infection.

4.Retail and Commercial Establishments:

• Integrate the system in shopping malls, supermarkets, and retail stores to enforce
mask usage and create a safer shopping environment for customers and staff.

4. Project Planning
The project planning should be given using either Gantt Chart or any other methodology
showing the clear picture of the progress of the project along with the involvement of the
group members in each activity of the project.

References

[1]. Viola, P., & Jones, M. J. (2004). Robust real-time face detection. International Journal
of Computer Vision, 57(2), 137-154.
[2]. Zhang, K., Zhang, Z., Li, Z., & Qiao, Y. (2016). Joint face detection and alignment
using multitask cascaded convolutional networks. IEEE Signal Processing Letters, 23(10),
1499-1503.

[3]. Simonyan, K., & Zisserman, A. (2014). Very deep convolutional networks for large-
scale image recognition. arXiv preprint arXiv:1409.1556.

[4]. Chollet, F. (2017). Xception: Deep learning with depthwise separable convolutions. In
Proceedings of the IEEE conference on computer vision and pattern recognition (CVPR)
(pp. 1251-1258).

[5]. Official OpenCV Documentation: https://docs.opencv.org/4.x/index.html

[6]. Rosebrock, A. (2018). Practical Python and OpenCV: An Introductory, Example-


Driven Guide to the Fundamentals of Computer Vision and Image Processing.
PyImageSearch.

[7]. TensorFlow Documentation: https://www.tensorflow.org/

[8]. Kaggle Datasets: https://www.kaggle.com/datasets

[9]. Prajna Bhandary, U., Zhang, L., & Lu, J. (2020). Bacterial foraging optimization
algorithm based deep transfer learning of VGG-16 for face mask detection in the era of the
COVID-19 pandemic. Computers, Materials & Continua, 63(2), 537-551.

[10].OpenCV-Python.Tutorial.2021.GeeksforGeeks.Available.at:
https://www.geeksforgeeks.org/opencv-python-tutorial/.
Guidelines for preparing Minor Project Synopsis:

Paper/ Page Size


Good quality only white A4 size (210 mm by 297 mm, 8.26 by 11.69) paper should be
used for typing and duplication.

Page Specification
Left margin: 1.25 inch
Right margin: 1 inch
Top margin: 1 inch
Bottom margin: 1 inch

Page Numbers
All text pages as well as program source code listing should be numbered at the bottom
right of the pages.

Font size
Normal Body Text: 12-point type size, Times New Roman
Section Heading: 14-point type size, Times New Roman, Bold
Subsection Heading: 12-point type size, Times New Roman, Bold & Italics
Line Spacing: 1.25 spacing.

Justification
The whole document should be both sided fully justified.

Headers and Footers


Headers and footers should be provided in the document, if required.

Tables and Graphs


Minimum 12-point type size, minimum line thickness 0.13 or .30 mm, all captions should
be upper-and lower-case, bold, and centred.

Illustrations and Photographs


Minimum 12-point type size, captions should be in upper-and lower-case, bold, and
centred. Images must be computer-designed and submitted as embedded images in your
document (postscript, .pdf, or MS Word format). Please do not submit colour images.

Binding and Color code of the Minor Project Synopsis


Minor Project Synopsis should be spirally bound with front cover being transparent and
back cover being Gray.

You might also like