The Design And Research Of The Auxiliary Navigation System Based On Augmented Reality

Augmented reality is a new technology developed on the basis of the virtual reality in the 1990 s.In AR,virtual information is superimposed onto the target’s vision in the real world,so as to render and enhance the interested target visually.Moreover,the target can be sensed in real time via a voice interface.In both ways,the situation awareness capabilities are significantly enhanced for the users.In this paper,a reverse projection transformation is proposed to solve homography matrix,and then a vehicle aided navigation system is established using AR. The following work is done in this paper:1. According to the characteristics of the structured road,a linear model is established.Furthermore, a series of pre-procession operations have been done on the original road image.Finally, lane detection and recognition is achieved by the algorithm of Hough Line Detection.2. In terms of the lane enhancement, a method based on reverse projection transformation to solve homography matrix is proposed relying on the analysis of previous homography matrix solutions employed in the marker-based and marker-less augmented reality systems.Taking advantage of the output homography matrix, virtual object superposition and registration are realized,which results in a genuinely new way for a driver to make some sense of the world around it and effectively prevent the phenomenon of deviating from the pathway caused by the inattention of drivers happenning.3. The traffic sign is divided into four categories: warning signs, prohibited signs, directional signs and direction signs by comparative analysis for traffic signs.In this paper, different traffic signs are detected by the chromatic aberration method based on the color feature.After detecting the underlying traffic signs,we extract their external contours according to the connectivity analysis.In this way,the traffic signs are detected and extracted.4. The Hu invariant moment feature is used as the input of the BP neural network to classify the traffic signs since different traffic signs use different shapes for easy identification and different shapes have different Hu invariant moment.And then, traffic signs are recognized based on their background colors,which can effectively reduce the loss rate of the traffic signs.Finally, our implemented system has realized the function of sound alarm.