Human eye fatigue detecting method

Abstract

The invention discloses a human eye fatigue detecting method. The human eye fatigue detecting method includes: outputting a final human face area via human face detection by a human face detector and a skin color model; realizing human eye positioning on the final human face area via a human eye detector and a human eye compositing template, outputting the precisely positioned human eye area, and updating the human eye compositing template; subjecting the precisely positioned human eye area to human eye outline point sampling so as to obtain sampling points, subjecting the sampling points to ellipse fitting, judging human eyes to be closed when ratios between major semi-axes and minor semi-axes are smaller than a second threshold value, or judging the human eyes to be open; and judging the human eyes to be in fatigue state when the human eyes are in closed state in continuous N frames or occupation of the closed state within a threshold time is larger than the threshold value ratios, or judging the human eyes to be in waking state. The human eye fatigue detecting method is less prone to effect of individual factors and light rays, is strong in robustness, instantaneity and accuracy, and effectively lowers the possibility of misinformation.

Description

technical field [0001] The invention relates to the fields of image processing and pattern recognition, in particular to a method for detecting human eyestrain. Background technique [0002] Fatigue is a subjective feeling of discomfort, but objectively, under the same conditions, it will lose the ability to complete the normal activities or work previously engaged in. Therefore, in some special positions, fatigue will affect the quality of work and the completion of tasks, and even cause inevitable disasters. For example: driving fatigue will affect the driver's attention, feeling, perception, thinking, judgment, will, decision and movement. Fatigue driving may cause unsafe factors such as delay or premature action, operation pause or improper correction time. , Highly prone to road traffic accidents, therefore, it is strictly forbidden to drive a vehicle after fatigue. In the field of national security, fatigue is also an important hidden danger that leads to a decrease ...

AI Technical Summary

Problems solved by technology

[0004] 1) Detection method for physiological signals: real-time detection of physiological signals of the human body by means of medical instruments and equipment, so as to judge whether the human body is fatigued; this detection method needs to use relatively complicated contact devices in application, which is not only expensive but Due to the limitation of the device, its development space is weakened;

[0005] 2) The detection method for the vehicle operation status: monitor the vehicle driving route, vehicle running speed, relative lane deviation degree, etc. through sensors, and then analyze whether the vehicle is running normally and whether the driver's operation is reasonable, and judge the driver based on this Fatigue state; this detection method is greatly affected by external factors such as vehicle type, road conditions and weather, and has limited application fields, so it can only be applied to driving fatigue detection

[0008] Existing detection methods for human body response characteristics are easily affected by light and individual factors, resulting in low detection accuracy; and poor real-time performance, there are false positives, affecting work quality and task completion

Images