The G drive link below contains the presentation desk, video presentation, model and the data that has been used to train the AI.
https://drive.google.com/drive/folders/14jBTddCsa6_qHcjO8KjS8K11wafrO-OV?usp=drive_link
Introduction- This GitHub repository houses the code and resources for a Drowsiness Detection System and Virtual Bumper for road vehicles. These advanced safety features aim to enhance driver alertness, reduce the risk of accidents, and prevent speed-related collisions. By implementing these technologies, we aim to improve road safety, minimize accidents, and reduce associated legal and insurance costs.
Drowsiness Detection System- A drowsiness detection system is designed to monitor driver behavior, typically through cameras or sensors, in order to identify signs of fatigue. Here are some key benefits of implementing such a system:
- Reducing the risk of accidents caused by drowsy driving.
- Minimizing fatalities and injuries, leading to fewer legal and insurance costs for businesses.
- Increasing productivity and operational efficiency by ensuring driver alertness in commercial fleets.
- Enhancing brand reputation and customer trust through a commitment to safety.
- Potentially reducing insurance premiums for businesses with safer driving records.
Virtual Bumper- A virtual bumper is a safety feature that uses sensors and algorithms to detect instances of error during driving and autonomously applies brakes or reduces speed to avoid accidents. Here are the advantages of implementing a virtual bumper system:
- Actively monitors and manages vehicle speeds, reducing the likelihood of high-speed accidents.
- Swiftly detects speed-related collisions, minimizing the severity of impacts and decreasing the risk of fatalities and severe injuries.
- Fewer speed-related accidents mean lower legal and insurance costs for businesses, leading to significant cost savings.
- Demonstrates a commitment to preventing speed-related accidents, enhancing a company's reputation for safety and responsibility.
- Ensures the system's effectiveness in speed-related accident prevention, supporting uninterrupted business operations and long-term sustainability.
Technology Stack- The Drowsiness Detection System and Virtual Bumper utilize the following technologies and tools:
OpenCV: Used for gathering images from webcams, which are essential for monitoring driver behavior.
AI Data Feeding and Training: A dataset of 2900+ images of people's eyes under different lighting conditions is used for training the Drowsiness Detection System.
Convolutional Neural Network (CNN): The CNN architecture is employed, consisting of input, output, and hidden layers, for accurate drowsiness detection.
Python: Knowledge of Python, including its basic syntax and structure, is required for working with the code.
TensorFlow: Used for the backend of the project, providing the framework for implementing neural networks.
Keras: Utilized for the classification aspect of the project, enabling efficient model training and evaluation.
Pygame: Utilized for playing alarm noises when drowsiness is detected, alerting the driver.
EGM (Ephemeral Grid Mapping): Employed for the virtual bumper functionality, enhancing collision avoidance capabilities.
Prerequisites Before using this project, ensure you have the following prerequisites:
● Knowledge of Python, including its basic syntax and structure. ● Installation of TensorFlow for the backend of the project. ● Installation of Keras for classification purposes. ● Pygame for playing alarm sounds. ● EGM (Ephemeral Grid Mapping) for virtual bumper functionality.