Being alert at the wheel, is a pressing problem concerning road safety, that many people face in their daily life. This project proposes developing technologies that can help to detect how alert the driver is. It addresses seven different use cases, to provide better feedback to the driver and therefore helps in reducing the number of accidents. The prototype performs Emotion Analysis through facial expressions, Visual Analysis of traffic signboards, calculates the driver’s Eye Aspect Ratio to detect drowsiness and mobile distractions and also calculates object proximity from the car using techniques of Computer Vision and Neural Networks. Drunk driving is detected using an alcohol sensor module MQ3. It also provides monthly alerts to the car owner about his vehicle maintenance and service. Every use case has a unique response which aims at alerting the driver to tackle the vital issue of safety at the wheel.
Driver Alertness System using Deep Learning, MQ3 and Computer Vision
2020-05-01
1645847 byte
Conference paper
Electronic Resource
English
IEEE | 2006
|SAE Technical Papers | 2024
|Vision-Based Driver Authentication and Alertness Detection Using HOG Feature Descriptor
Springer Verlag | 2020
|