Research On The Statistical Distribution Of The Driver Head Characteristics Points Based On Binocular Vision Measurement

The distribution of driver head characteristics points is very important to the vehicle package design and passive safety.However few corresponding research exists in China,especially one based on the Chinese anthropometry,which make the condition that no suitable reference data can be used when developing vehicles facing on the Chinese population.And because of the difference in anthropometry of Chinese and other population,foreign standards are also not suitable for Chinese population.Therefore,in this thesis the statistical distribution of the driver head characteristics points is studied by laboratory test based on the binocular vision measurement.The main contents of this thesis are:1)Studied the binocular vision measurement method,including the camera imaging model,mathematical model of binocular vision measurement,camera calibration,image pre-processing,extraction of the pixel coordinates,pixel points matching and 3D coordinates calculation and conversion.In the pixel coordinates extraction,the grayscale weighting mass center method is used to ensure that the precision is at the Sub-pixel level.A cubic matching method is established which is suitable to be used in this thesis.2)A binocular vision measurement system is constructed,including the hardware construction and software programming.The cameras are selected and debugged.Calibration plates are made.A camera bracket is designed and made which can be used for different snapshot distances and backgrounds.Software for image acquiring,camera calibration,extraction of pixel coordinated,3D coordinates calculation are programmed.3)The experiment is designed.By analyzing the Chinese anthropometry data,the human body size information is acquired,which is used as the basis for participant selection to ensure that the anthropometry can cover the target population.The vehicle data is selected.The test strategy is made to ensure the accuracy of the data.4)The top,eye,headrest and Tragion point are selected as the research object.Themeasurement test is conducted to acquire these points of the participants,by using the binocular vision measurement system.The participants are at the normal driving posture,and their head characteristics points are measured.The coordinates of the head characteristics points are acquired and converted.5)The data is statistically analyzed,and the prediction models are established for the participants' head characteristics points.The distribution of these head characteristics points can be predicted using the body size and posture.By above test and statistics analysis,the function of the binocular vision measurement system is validated.The prediction models of the distribution of the driver head characteristics points at normal driving posture are acquired,which can be used to help the cab packaging design and seat headrest design.