Research On Vehicle-based HUD Human-Computer Interaction System Based On Gesture Recognition

According to statistics,the traffic accident caused by driver distraction accounts for 65%,and the driving distraction is mainly caused by the driver looking down at the dashboard and operating related equipment.In this paper,the gesture recognition technology is applied to the vehicle HUD to design a human-computer interaction system suitable for the interior environment.Optimize the driving experience while ensuring driving safety.Aiming at the problems of static gesture recognition that are susceptible to skin-like objects and background interferers.The gesture segmentation combined with skin color segmentation and background difference method is proposed,and the skin color and background model are updated in real time.The gesture characteristics are obtained by the Hu moment and the convexity of the gesture contour.The support vector machine is used to classify the gesture features.Experiments show that the average recognition rate of the two groups of static gesture databases is 99.05%.For the dynamic gesture recognition,the real-time performance is poor and the gesture is easy to lose.Firstly,the optical flow method is used to calculate the corner optical flow vector in the key frame image for segmentation of the gesture region.The Kalman filter is combined with the optical flow method to predict the position of the tracking gesture in real time.The gesture center point is extracted to extract the gesture trajectory feature.The dynamic time warping algorithm classifies.Experiments show that the average recognition rate of the algorithm in the dynamic gesture database is 97.87%.The gesture recognition module is mounted on the vehicle HUD platform,and the system simulation test is performed in the vehicle.Using the algorithm proposed in this paper,the success rate of gesture recognition reaches 95.17%,and 16 functions are commonly used in the car to meet the needs of human-computer interaction.