Drunk driving detection method and system based on sensor and machine vision

Abstract

The invention belongs to the technical field of drunk driving detection, and particularly relates to a drunk driving detection method based on a sensor and machine vision, which comprises the following steps: collecting a training sample: respectively obtaining the alcohol concentration in a vehicle, the heart rate of a driver, the body temperature and a face image by utilizing an alcohol concentration sensor, a heart rate sensor, a body temperature sensor and a camera which are arranged in the vehicle; obtaining a drunk driving identification result of the image: performing face positioning and drunk driving identification on the face image by using an MTCNN network and a VGG16 network; establishing a BP neural network model: carrying out BP neural network training on the alcohol concentration in the vehicle, the heart rate and body temperature of the driver and the drunk driving recognition result of the image in the training sample; and carrying out drunk driving identification: inputting the alcohol concentration in the vehicle, the heart rate of the driver, the body temperature and the face image acquired in real time into a BP neural network model to judge whether the driveris in a drunk driving state. The drunk driving identification method can effectively improve the drunk driving identification rate.

Description

technical field [0001] The invention belongs to the technical field of drunk driving detection, in particular to a method and system for detecting drunk driving based on sensors and machine vision. Background technique [0002] With the continuous development of modern society and science, the means of transportation are also constantly improving. In daily travel, automobiles have become the main means of transportation for people; although the advancement of transportation has brought a lot of convenience to people's life and work, frequent traffic accidents have caused great harm to people's lives and properties. threaten. According to the World Health Organization survey, about 50-60% of traffic accidents are related to drunk driving, so drunk driving identification is of great significance to prevent drunk driving. [0003] At present, the monitoring method of drunk driving behavior is mainly manual monitoring. The traffic police use a portable alcohol detector to dete...

AI Technical Summary

Problems solved by technology

The third-party remote video is actually human supervision and subjective judgment, which makes the final result have subjective errors and reduces the accuracy of recognition

However, this method mainly relies on the accuracy of alcohol concentration detection. Video recognition only detects whether there is a substitute phenomenon in the driver instead of performing image recognition to judge whether the driver in the image has drunk driving behavior. When the alcohol sensor is artificially blocked, the drunk driver is driving. If there is no drinking behavior in the process, it cannot be identified, and the patent document does not give the method used for video recognition

[0007] The application number is 201611267904.6, which discloses a real-time monitoring method of dangerous driving behavior based on deep learning. Through the image collection of the preceding vehicle, a data set of dangerous driving behavior is established, and the recognition model of dangerous driving behavior is obtained by using deep learning training. Although the dangerous driving behavior in the patent does not include Drunk driving is not included, but the facial features of the driver after drinking and driving are also different from the normal state, for example, the face turns red, the eyes become smaller than usual due to blurred consciousness, and the mouth slightly opens, etc.

Images