TAMILNADU STATE COUNCIL FOR SCIENCE AND TECHNOLOGY

STUDENT PROJECT PROPOSAL

1. Name of the Student (s) :

[Link] Name of the Student E-Mail ID Phone No.

1 V. Subramanian 953622205048@[Link] 9751375567

2 G. Dinesh Pandian 953622205017@[Link] 8148411101

3 G. Ram Pandian 953622205034@[Link] 6382580509

4 S. Ragul Raj 953622205033@[Link] 81486 47696

2. Name of the Guide : Dr. K. Palraj

Department / Designation : Associate Professor,

Department of Information Technology,

Institutional Address : Ramco Institute of Technology,
Rajapalayam, Virdhunagar District

Phone No. & Mobile No.

: 8508259397

3. Project Title : Next-Generation Intelligent Water Distribution and

Quality Controlling system
4. Sector in which your Project proposal is to be Considered : Engineering Technology (Computer Science)

5. Project Details : (write up (max. 3 pages only) should be given for each item including justification for acquiring
/ fabricating equipment’s / apparatus in the budget)

i. Introduction iv. Work Plan

ii. Objectives v. Budget

iii. Methodology vi. Any other details Enclosed

6. Has a similar project been carried out in your college / : No

elsewhere? If so furnish details of the previous
project and highlight the improvements suggested
in the present one

CERTIFICATE
This is to certify that Mr. V Subramanian, Mr. G Ram Pandian, Mr. G Dinesh Pandian, Mr. S Ragul Raj is a bonafide Pre-final year student of U.G. Engineering courses of our

college and it is also certified that two copies of utilization certificate and final report along with seminar paper will be sent to the Council after completion of the project by the end

of May 2025.

Signature of the Guide Signature of the HOD Signature of the Principal

Head of the Institution
(with seal)

Real Time Sign Language Detection with Tensorflow Object Detection

and Python|Deep Learning SSD

 INTRODUCTION

The Real-Time Sign Language Detection system leverages TensorFlow's object detection framework and advanced deep learning techniques, specifically the SSD (Single Shot

MultiBox Detector) model, to accurately identify and interpret sign language gestures in [Link] system integrates TensorFlow's pre-trained SSD model to identify and

classify sign language gestures from live video feeds. By leveraging the power of deep learning and efficient object detection, the system can recognize various sign language

gestures in real-time, providing immediate feedback and facilitating communication. The application is implemented in Python, utilizing libraries such as OpenCV for video

capture and TensorFlow for object detection, to deliver an effective solution for enhancing accessibility and communication for the hearing-impaired community

i) OBJECTIVES

The main objectives of this project are

 Accurate Gesture Recognition: Develop a system that precisely identifies and interprets various sign language gestures from live video feeds.

 Real-Time Processing: Ensure the system delivers immediate feedback by processing video input in real time.

 Integration of TensorFlow and SSD: Utilize TensorFlow’s object detection framework and the SSD model for efficient and effective gesture detection.

 Enhancing Accessibility: Provide a tool that translates sign language gestures into understandable formats, improving communication for the hearing-

impaired.

 Efficient Resource Utilization: Optimize the system for performance on standard hardware to ensure broad accessibility.

 Robust Performance: Maintain reliable detection across varying lighting conditions and backgrounds for consistent results.

ii) METHODOLOGY

The methodology for Real-Time Sign Language Detection using TensorFlow's Object Detection and SSD Model involves several key steps: First, gather and annotate

a diverse dataset of sign language gestures. Next, select and fine-tune a pre-trained SSD model from TensorFlow's model zoo to adapt it for gesture detection.

Implement the system by integrating video capture using OpenCV, preprocessing video frames to fit the model's input requirements, and running real-time object

detection to identify and classify gestures. Post-process the detection results to display bounding boxes and labels on the video feed, and ensure the system is

efficient and accurate through rigorous testing and optimization. Finally, deploy the system on target devices and integrate it with applications to enhance

communication accessibility.

iii) Block Diagram:

iv) WORK PLAN

The project begins with defining objectives, gathering requirements, and assembling the team, followed by the procurement of sensors and hardware, and the setup of a

real-time monitoring system for water quality and distribution. We develop data processing algorithms and automated alert mechanisms, alongside a user-friendly

dashboard for displaying real-time data, historical trends, and alerts.

After thorough testing in both simulated and real-world environments, we integrate user training and feedback to refine the system. The final phase focuses on

optimizing performance through regular calibration, software updates, and overall system efficiency, culminating in the full deployment, evaluation, and impact

assessment to ensure effective and reliable water management.

Month (2025)

Work January February March April

1-15 16-31 1-15 16-29 1-15 16-31 1-30

 Study of Proposed System and Relevant Water Quality Standards

Study and Selection of Sensor and Communication Technologies

Development of Real-Time Monitoring Software and Dashboard

Testing, Report Preparation and System Deployment

v) BUDGET

AMOUNT
[Link] EQUIPEMENT QTY JUSTIFICATION
(in Rs.)

1 Sensors and Components for the project 1 8000 -

3 Miscellaneous - 2000 Travelling charges, data collection, etc.,

Total 10000 -

Total Budget Amount: Rs. 10,000/-