A method for automatically detecting the driver's observation state

Abstract

The invention discloses a method for automatically detecting observing states of a driver. The method comprises the following steps of detecting the space attitudes of a driver and a motor vehicle through three-axis gyroscopes which are arranged on a head of the driver and the motor vehicle body, and calculating to obtain the motion attitude of the driver corresponding to the motor vehicle, so as to obtain the observing state of the driver in the motor vehicle driving process; in combination with the current examination item and evaluation standard, automatically evaluating and judging the examination performance of the driver, so as to realize the automation effect of road driving skill examination of the driver. The method has the advantages that the structure is simple, the design is skillful, and the convenience in use is realized.

Description

technical field [0001] The invention relates to the technical field of a road driving skill test system of a motor vehicle driving management department, in particular to a method for automatically detecting a driver's observation state. Background technique [0002] According to the latest requirements of the Ministry of Public Security's "Regulations on the Application and Use of Motor Vehicle Driving Licenses", in the process of applying for a motor vehicle driving license, applicants must pass the examination subjects of road traffic safety laws, regulations and related knowledge (referred to as "Subject 1") 1. Test subjects of venue driving skills (referred to as "subject 2"), test subjects of road driving skills and general knowledge of safe and civilized driving (referred to as "subject 3"). The requirements for the examination and evaluation of each subject are clearly stated in the regulations. [0003] In the road driving skills (subject 3) test subject, key asses...

AI Technical Summary

Problems solved by technology

[0004] (1) If the line of sight leaves the driving direction for more than 2 seconds, it is unqualified;

[0005] (2) Those who fail to observe the interior and exterior rearview mirrors, and fail to observe the traffic conditions behind them before starting, are unqualified;

[0006] (3) When driving in a straight line, 10 points will be deducted for observing the rear traffic conditions through the interior and exterior rearview mirrors inappropriately;

[0007] (4) Before changing lanes, failing to observe the rear road traffic conditions through the interior and exterior rearview mirrors, the vehicle is unqualified;

[0008] (5) Failure to observe the traffic situation behind and on the right side through the interior and exterior rearview mirrors before parking is unqualified;

[0009] (6) After parking the car, failing to observe the traffic conditions on the rear side and left side before opening the door inside the car is unqualified;

[0010] (7) When going straight at an intersection, turning left at an intersection, or turning right at an intersection, failing to observe the traffic conditions on the left and right, and failing to observe the traffic conditions in front of the side when turning through the intersection, the driver is unqualified;

Using the method of manual judgment by the examiners, in the case of insufficient police force (many test centers in China only have 1 to 2 examiners), these assessment requirements are basically ignored, resulting in lower difficulty of the test and failing to screen qualified drivers the goal of

The method of using computer video image processing technology to detect the driver's observation state is to use the camera to collect the driver's face video stream in real time, and send it to the computer for processing. The computer determines the driver's face (or eyes) by processing the input video image. ) features, and then determine the driver’s line of sight direction, that is, determine the driver’s observation state; this method is greatly affected by changes in external light, and is prone to misjudgment, resulting in low detection accuracy and difficult to be used in actual tests

Images