Project Proposal: Road Sign Detection System for Harsh Driving Analysis and
Speed Compliance
1. Problem Statement:
Harsh driving, characterized by sudden acceleration, hard braking, and abrupt lane
changes, poses a significant risk to road safety, highlighting the need for a robust
system that can accurately detect and analyse such behaviours by combining multiple
data sources, including vehicle speed and visual cues from the surrounding
environment.
2. Proposed Solution:
The proposed solution is to develop a real-time system that integrates data from
Inertial Measurement Units (IMU), Global Positioning Systems (GPS), and camera
inputs to accurately detect and analyse instances of harsh driving. By combining
vehicle dynamics data, such as speed and acceleration, with visual cues from the
surrounding environment, the system will provide a comprehensive approach to
identifying and evaluating rash driving behaviours, offering actionable insights to
enhance road safety.
3. Objective:
The objective of this project is to design and develop a real-time system that
integrates data from IMU, GPS, and camera inputs to accurately detect, analyse, and
evaluate instances of harsh driving, with the goal of enhancing road safety through
comprehensive monitoring and actionable insights.
4. Hardware Requirements:
The initial list of Hardware required for prototyping of the concept includes:
1. Microcontroller capable of executing multiple codes simultaneously
2. Camera to Capture surroundings
3. Secondary sensors such as Accelerometer and GPS for additional data.
4. Hardware accelerator to meet computation requirements.
5. Devices to show output of the system.
6. Power source.
7. Data storage device.
8. Cooling systems to reduce system temperature.
5. Software Requirements:
1. Open-source ML model to train on custom dataset.
2. Operating system able to run on microcontroller.
6. Project Flow:
7. Budget Breakdown:
Sr. No Component Price
1 Raspberry Pi 4 Model B/ 5 (8GB Ram) 7780
2 5MP Raspberry Pi Camera 220
3 Heat Sink/ Cooling Fan 485
4 MPU6050 Accelerometer 250
5 Google Coral TPU 10600
6 Battery 3000
7 Raspberry Pi Micro SD Card 64GB/ 128GB 1100
8 Raspberry Pi Case with Fan 990
9 GPS Neo8M Sensor 650
10 LM2596 DC-DC Adjustable Converter 500
11 USB to Type-C Cable 200
Total 25775
8. Expected Outcome:
The project defined according to its scope must give the following outcomes:
1. It should be able to classify Traffic signal into 3 categories namely Red, Yellow and
Green.
2. It should be able to identify ‘numbers’ present in the speed limit sign.
3. The Buzzer must sound when the rules defined in the code are violated. For eg:
For speed limit of 60, the speed of the vehicle must not be above 60 for that
particular time period.
9. Conclusion:
The proposed system aims to address the critical issue of harsh driving by leveraging
the integration of IMU, GPS, and camera inputs for real-time detection and analysis.
By combining vehicle dynamics data with visual cues, the system provides a
comprehensive solution to identify and evaluate rash driving behaviours, contributing
to improved road safety and enabling proactive measures to mitigate risks associated
with unsafe driving practices.
10. Group Details:
Sr. No Name Roll No
1 Saurabh Autade BEN 04
2 Nishad Harsulkar BEN 14
3 Sujal Koli BEN 20