Research On Eye Tracking System Of Computer Optometry Instrument

With the growth of myopia population and the development of science and technology,people need more automated computer refractometer. At present, common computerrefractometer mostly only own semi-automatic eye tracking system, which meansoptometrist’s manually adjusting and instrument automatic tracking have to be combined.Eyeball positioning make the process to be low speed, poor accuracy and a relatively lowlevel of automation.After analyzing current tracking technology and the characteristics of refractometer, aneye tracking system based on photoelectric tracking technology is designed in the paper.The eyeball tracking system is fully automatic, and eye tracking can be done without anyoptometrist operation. The epigynous machine and the hypogynous machine of the systemare completed in the paper. The epigynous machine part mainly includes video imageacquisition module and video image processing module,etc. The hypogynous machine partmainly includes hardware control module and three axis stepper motor actuators.’Video image acquisition module is used to get the data by means of active light visualmethod, using coaxial infrared auxiliary light source to illuminate the target area. ThenCCD is used to receive corneal reflex Infrared light spot imaging from the target area, andthe data will be transmitted to video image processing module through the API interface.Video image processing module calculates the parameters of miss-distance of controltracking and sharpness evaluation of control focusing. Image preprocessing part,identification, parameter calculation consists the module.The hypogynous machine, FPGA control board and dedicated three axis motor driverboard are designed and completed through combined with tracking, focusing algorithm andstepper motor working principle. FPGA control board receives instructions through a serialport, and drives actuators by control the motor drive board, converting the angular displacement of stepper motor to linear displacement of a ball screw and the precisepositioning of screw is realized, thus implementing high precision tracking function of thesystem.By testing and analyzing the results of epigynous machine and lower position machine,the image processing algorithm, tracking focus algorithm and hardware control strategy ofthe human eye tracking system are verified to be feasible. The system possesses thecharacteristics of nice tracking effect and very high tracking accuracy.The summary of work done are shown at the last of the paper, and the system needs tobe improved in the design of next stage was discussed.