Driver Drowsiness Detection System Using Image Processing
Driver Drowsiness Detection System Using Image Processing
ABSTRACT:
In the realm of road safety, the “Driver Drowsiness Detection System Using Image Processing” emerges as a pioneering technological advancement. This project harnesses the power of MATLAB, employing a strategic blend of image processing techniques and machine learning algorithms – K-Nearest Neighbors (KNN) and Random Forest – to achieve an impressive accuracy rate of 97%. The proposed system offers a comprehensive solution, encompassing image acquisition, pre-processing, segmentation, feature extraction, and classification stages.
In the initial phase of image acquisition, data is collected from the vehicle’s interior environment, providing crucial input for subsequent analysis. The subsequent pre-processing stage plays a pivotal role, commencing with grayscale conversion to enhance computational efficiency. Moreover, the system employs advanced techniques for eye detection, ensuring precise localization of the driver’s eyes within the captured images.
The following phase of the system is dedicated to segmentation, takes center stage by executing intricate IRIS segmentation and extraction processes. This critical step serves as the foundation for accurate feature extraction in the subsequent module. The next phase employs Discrete Cosine Transform (DCT) and Speeded-Up Robust Features (SURF) to extract discriminative features from the segmented iris images, facilitating robust identification of drowsiness-related patterns.
In the final phase, the system enters the classification phase, utilizing both KNN and Random Forest classifiers. These machine learning models have been fine-tuned to deliver exceptional accuracy in distinguishing between alert and drowsy states. Additionally, the project introduces an innovative concept – the Fusion score, calculated as the weighted average of KNN and Random Forest scores, utilizing the formula KNN+RF/2. This fusion mechanism enhances the system’s reliability by leveraging the strengths of both classifiers.
The culmination of these efforts results in the system’s ability to make a decisive determination: whether the driver’s eyes are open and alert or exhibiting signs of drowsiness. The seamless integration of image processing, machine learning, and fusion techniques makes this project an indispensable tool in enhancing road safety by preventing accidents caused by driver drowsiness.
In essence, the “Driver Drowsiness Detection System Using Image Processing” showcases the tremendous potential of MATLAB, KNN, and Random Forest in the realm of road safety. With an exceptional accuracy rate of 97%, it stands as a testament to innovation and engineering prowess, offering a lifeline to countless lives on the road.
PROJECT OUTPUT VIDEO:
ALGORITHM / MODEL USED:
K-Nearest Neighbors (KNN) and Random Forest.
SYSTEM REQUIREMENTS:
HARDWARE REQUIREMENTS:
- System : Pentium i3 Processor.
- Hard Disk : 500 GB.
- Monitor : 15’’ LED.
- Input Devices : Keyboard, Mouse.
- Ram : 8 GB.
SOFTWARE REQUIREMENTS:
- Operating system : Windows 10 Pro.
- Coding Language : MATLAB
- Tool : MATLAB R2023b
REFERENCE:
Duy-Linh Nguyen , Member, IEEE, Muhamad Dwisnanto Putro , Member, IEEE, and Kang-Hyun Jo , Senior Member, IEEE, “Lightweight CNN-Based Driver Eye Status Surveillance for Smart Vehicles”, IEEE Transactions on Industrial Informatics (Early Access), 2023.
