Research And Application Of Navigation Method Based On Augmented Reality

Augmented Reality(AR)is a kind of technology that can integrate the real scene and virtual information in real time.With the development and improvement of augmented reality technology,more and more applications of augmented reality have been promoted to get more people’s attention and interest.The current augmented reality is mainly achieved on special hardware devices.But not only this,the processing capability of the mobile phone has been greatly improved,that can satisfy the small augmented reality application.To enhance the integration of reality and navigation information and the integration of virtual navigation directive information fusion and the real world,giving the user a better navigation experience with the information more clearly,easier and more direct.In the augmented reality technology,the important algorithm is tracking registration algorithm which is how to combine the real world and virtual information According to the above characteristics,the algorithm required to ensure the accuracy of the algorithm,while reducing the amount of computation and improve the speed of the algorithm.In this paper,aim at analyzing and studying the application of the vehicle navigation system in the outdoors.The ideal effect is that when the users travel,the 3D navigation model can be rendered in real scene in real time to indicate the user’s direction of travel.Through the study and comparison of the main tracking technology of augmented reality,this paper puts forward the method of tracking and registering more suitable for using in outdoor vehicle navigation.Because in the driving process,the scene changes quickly,and cannot be used to match the two key points to obtain the relationship between the camera and the virtual model.But in order to ensure the registration of virtual navigation marks indicating the same direction of the road,choose to obtain road information from the vision,in order to get the accurate registration information.In terms of visual processing,this paper selects the direction information of common information Lane on the road as a virtual object registration.In the visual processing,using the lane recognition method based on the detecting the vanishing point of lane,firstly Y value and Cr value are used to cluster the color information of the image,and then detect the vanishing point in the image after segmentation,thus completing the recognition of lane.In order to increase the smoothness of identification and avoid the impact of jitter and bump on lane identification,the Calman filtering method is used to predict and optimize lane.Through experiments,this paper in different scenarios of lane recognition and tracking have good effect in the virtual navigation mark registration,registration direction is accurate.The overall algorithm can satisfy the real-time identification of lane tracking,navigation mark will be registered for real-time display to the user’s requirements.