Opticam

Introduction

Opticam addresses critical road safety concerns, such as drowsy driving and car theft, which are significant contributors to vehicle accidents and theft worldwide. Many existing solutions lack real-time monitoring and advanced AI integration, leaving substantial gaps in safety and usability. Opticam aims to bridge this gap by introducing an innovative AI-powered dual-view dashcam that enhances driver alertness, provides theft protection, and improves overall road safety.

Key Features

  1. AI Drowsiness Detection: Utilizes YOLOv8 and custom-trained machine learning models to detect driver drowsiness in real time.
  2. Dual-View Monitoring: Records both the driver and the vehicle's surroundings for comprehensive coverage.
  3. Real-Time Alerts: Features customizable auditory and visual alerts to prompt driver awareness and prevent accidents.
  4. Mobile App Integration: Includes a user-friendly application designed with Figma for seamless interaction, allowing users to manage devices, receive notifications, and share footage.
  5. Data Privacy and Security: Incorporates local storage and encrypted data handling to ensure user privacy.

Development and Innovation

Opticam employs advanced technologies, including YOLOv8 and Python, for real-time object detection and classification. The training process involved customizing datasets to improve detection accuracy and configuring various parameters such as batch size and resolution. A user-friendly interface was designed using Figma, focusing on accessibility and visual appeal. Additionally, algorithms for face detection and zooming were implemented to enhance driver monitoring. Internal testing identified challenges, such as low-light performance and issues with detecting individuals wearing sunglasses, which are being addressed in future iterations.

Impact and Future Directions

In the next phase, Opticam aims to enhance model accuracy for better detection of driver expressions and alertness, ensuring improved safety. Planned features include a visual sound alarm to alert drowsy drivers and a night mode to optimize video capture in low-light conditions. For the mobile application, new functionalities such as a download option for saving playback footage and enhanced security measures like PIN codes and face identification will be introduced.

The dual-view dashcam with drowsiness detection and real-time connectivity has significant growth potential. Future extensions include integrating advanced safety features like ADAS, human detection, and object detection, as well as child detection to prevent incidents involving children left in parked cars. The solution can also be adapted for industries like EV taxis, where interior cameras and GPS integration can leverage AI to monitor drivers and send alerts to both drivers and central control systems, expanding its usability and market potential.

Researcher
Phasin Noomkan
Student
CMKL University
Phurich Amornnara
Student
CMKL University
Thanakit Thanasuwanditee
Student
CMKL University
Puttipong Srisuwantat
Student
CMKL University
Advisor
Dr. Boonyarit Changaival
Adjunct Faculty Member
CMKL University