The Virtual Message Board System Based On Image Matching For The Scenic Spot

Tourism is the important way of leisure, but a large number of tourists bring huge damage to the scenic spot as well as great wealth. We can see graffiti everywhere in the scenic spot; these behaviors cause serious destruction to the scenic spot.In order to solve the problem of tourists doodling in the scenic spot, this paper presents a virtual scenic spot message board system. This system is a software based on Android mobile phone, it simulates users doodling on the real object by drawing on the photos of objects, which not only meets the user’s demand of doodling, but also avoids the destruction to the scenic spot.This system needs to realize the two basic functions: leaving a graffito on an object and viewing other tourists graffito. Leaving a graffito on an object is as follows: when a user wants to leave a graffito on an object, he can use the system to photograph the object first, then use the system to brush on the photograph and upload it to the database. Tourists can also see other tourists graffito on this object after they photograph the object. When tourists photograph an object, the system should just bring graffiti on this object, and not bring any graffito on other objects, which is consistent with the actual scene of doodling on the real object.In order to only get graffiti to the object which users photograph of, we need to distinguish between different objects accurately, after analysis, we put forward the solution which both using the mobile location technology and image matching technology. We add location information for objects and the photos of objects, when searching in the database, we first screen objects according to the user’s location information, as a result of the mobile phone positioning accuracy problem, results of screening are often more than one, then we will use image matching technology to match these results with pictures the user uploading,after this, we can make sure that which object the user photograph of, then we get these graffiti of this object. In the above scenario, we focus on image matching technology, we first select the SIFT algorithm and the ORB algorithm for experiments, and then in order to speed up the algorithm, the feature point detection algorithm is combined with color histogram, finally we give the solution which has great accuracy and speed.On the realization of the system, this paper has done the requirement analysis, feasibility analysis and profile design according to the actual situation. In the Android clients, we have done the detailed design of drawing function, Camera function and getting location information function. In the server, we have done the detailed design of image matching modular and feature extraction modular. Finally, we has done a good realized realization of each function of the system requiring, the system was tested and the test results show that the system is practical.