Scholarlink

This chapter outlines the design and execution of an intelligent system for identifying intoxicated drivers through the integration of facial image analysis and breath alcohol detection. The device utilizes a Raspberry Pi as the core computing unit, incorporating an MQ3 gas sensor to measure the concentration of alcohol in exhaled breath and a camera module to record facial characteristics. A lightweight deep learning model using MobileNetV3 is used to assess visual symptoms of intoxication, while the sensor offers physiological validation. The system facilitates real-time surveillance and provides prompt notifications to mitigate the danger of impaired driving. Experimental validation in simulated car conditions confirms the efficacy and reliability of the proposed method for embedded driving safety applications.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Evaluating Driver Impairment in Automobiles via Facial Characteristics of Intoxicated Operators

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • Nguyen Vo Thanh Duy,
  • Bui Thien Phu,
  • Bui Tuan Huy,
  • Bui Nhat Tan,
  • Nguyen Van Bay,
  • Nguyen Trung Hau,
  • Nguyen Quoc Trung

摘要

This chapter outlines the design and execution of an intelligent system for identifying intoxicated drivers through the integration of facial image analysis and breath alcohol detection. The device utilizes a Raspberry Pi as the core computing unit, incorporating an MQ3 gas sensor to measure the concentration of alcohol in exhaled breath and a camera module to record facial characteristics. A lightweight deep learning model using MobileNetV3 is used to assess visual symptoms of intoxication, while the sensor offers physiological validation. The system facilitates real-time surveillance and provides prompt notifications to mitigate the danger of impaired driving. Experimental validation in simulated car conditions confirms the efficacy and reliability of the proposed method for embedded driving safety applications.