An Improved Mobile Augmented Reality System

With the development of mobile devices, especially the wearable mobile equipment, augmented reality gradually closer to people’s life. Augmented reality is on the basis of the virtual reality technology, The biggest characteristic is that it can combining virtual information with the real world. Augmented reality consist of the image recognition, target tracking, virtual information and other technologies, for the limited resources of mobile terminals, It can not provide rich information for user without cloud server. So mobile image recognition based on the cloud server is also one of the important technology of mobile augmented reality system.This paper deeply discusses the Android platform of mobile augmented reality system which based on the cloud image recognition, and through extract image saliency features improve the response speed of the image recognition module, Then we do the experiment and results analysis. By applying the optimized scheme, and apply augmented reality to electronic commerce system.First of all, this article simply introduces the research background and significance, and then we talked about the image recognition, virtual reality for mobile information tracking and the development of Android technology and so on. At last we apply augmented reality to electronic commerce system. In order to make the cloud servers able to handle huge amounts of image data, the system of the server will be based on the structure of inverted index model. Finally through the experiment We compared the number of SIFT feature points with the saliency SIFT feature points and the extracting time also recognition accuracy, results show that whit the reduction of the number of image feature points, the system identification accuracy of target image isn’t going down.