| Abstract [eng] |
The usage of moving objects location information becomes very popular nowadays. This location information can be used in many places: for emergency purposes, tracking cars with cargos, animal tracking, tourism, etc. The moving object’s positions can be detected using location techniques, such as GPS, GSM, etc. When detected moving object’s position is kept in Moving Object Databases (MODB). When object’s position is detected constantly we get a lot of positions, which represent the movement of particular object. We can call this movement a trajectory. The main problem is that these trajectories are too big to keep them in database, to update them or to make a search among them. That’s why there is a need for effective trajectory compression methods in order to make trajectories smaller without losing the most important object’s movement information. Different location techniques generate different types of trajectories, so each type of trajectory needs a different approach how to apply compression technique to it. The trajectory compression can be applied while generating trajectory or it can be done when the generation of trajectory is finished. There are presented different trajectory compression methods for different location techniques (GPS, GSM) in this work. These methods are incorporated into location data history restoration system in order to apply them for location data compression. In this work statistical methods for evaluation of location data compression methods effectiveness are presented. Statistical results for these methods prove that methods are reliable and anyone can get the same results when using these methods. |
