Blink Detection Method Based on Circle Fitting

Abstract

The invention discloses a blink detection method based on circle fitting, comprising steps: 1. Image collection: use image collection equipment to collect the face image of the subject under test, and collect the face image collected at each sampling moment Synchronously sent to the image processor; 2. Image processing: the image processor analyzes and processes the facial images received in each analysis and processing cycle according to the preset analysis and processing frequency f and in chronological order, the process As follows: 201. Analysis and processing of facial images received in the first analysis and processing cycle, including steps: synchronous storage of facial images, eye blink detection and counting of blink times; 202. Analysis and processing of facial images received in the next analysis and processing cycle ; 203, return to step 202. The method of the invention has simple steps, reasonable design, convenient implementation and good use effect, and can easily and quickly detect the blinking state of the subject through the circle fitting.

Description

technical field [0001] The invention belongs to the technical field of image processing, and in particular relates to a blink detection method based on circle fitting. Background technique [0002] According to statistics, fatigue driving accounts for a large proportion of traffic accidents, especially in highways, the proportion of traffic accidents caused by fatigue driving is much higher than that of ordinary roads, mainly including long-time driving, lack of sleep or poor quality, circadian rhythm, driver factors. In recent years, driver fatigue monitoring and even physiological and psychological state analysis have attracted widespread interest from researchers all over the world, and research on fatigue driving monitoring has achieved certain results. [0003] At present, there are mainly four types of fatigue driving detection methods: first, fatigue driving detection based on the driver's physiological signal, specifically, in the field of biological information, a ...

AI Technical Summary

Problems solved by technology

[0003] At present, there are mainly four types of fatigue driving detection methods: first, fatigue driving detection based on the driver's physiological signal, specifically, in the field of biological information, a detection device or electrode is installed on the corresponding part of the driver, and the driver's heart rate, EEG or EMG signals are monitored to judge the driver's state. This method has achieved good results in early research; The relationship between the signal power spectrum values ​​of different frequency bands of the electrical signal and the driver's subjective fatigue evaluation, the results show that: the subjective fatigue evaluation corresponds to the change of the power spectrum value in the EEG signal, and the ratio of the EEG power spectrum (α+θ) The larger the / β, the higher the fatigue level, but due to the cumbersome installation of the equipment, inconvenient wearing, and high cost, it cannot be widely promoted in practical applications; second, the fatigue driving detection based on the driver's operation behavior, Northeastern University's Wang Fei and others combined EEG recognition with vehicle handling characteristics to detect the driver's fatigue state, which is expected to provide theoretical and experimental basis for building a fatigue driving detection system, and designed a simulated driving experiment to collect the subjects' EEG Figure (EEG) signal and the corresponding steering wheel manipulation data; Evaluate the driver's fatigue degree according to the vehicle handling characteristics to determine the classification standard of the EEG signal, and select the support vector machine to classify the EEG signal to complete the qualitative analysis of the driver's mental fatigue state ; Due to the different operating habits of different drivers or irregular operating actions, the detection effect of this method is not ideal; the third, the detection method based on vehicle state information, the most representative is the AutoVue system developed by Iteris Corporation of the United States. It detects the driver's driving trajectory through a CCD camera facing the front of the road, and when the driver unconsciously deviates from the lane due to fatigue (such as the turn signal is not turned on), it can warn the driver in time; Zhong et al. use energy analysis and Wavelet analysis technology, through the monitoring of vehicle trajectory and steering wheel rotation angle, also realizes the detection of driver fatigue driving; but due to the complex real road conditions and complex changes in road conditions, the detection effect in the actual environment is poor; Fourth, based on the detection method of the driver's physiological response characteristics, this method can accurately identify the driver's state, but it is difficult to detect the driver's physiological response characteristics

For example, when monitoring the driving state according to the number of blinks of the driver, there is currently a lack of a method for detecting the blinking state of the driver with simple steps, reasonable design, easy implementation, and good use effect, which can easily and quickly detect the blinking state. And the accuracy of the test results is high

Images