Trabajo de grado Final (original) (raw)

Knowing the location of the nodes in wireless sensor networks is essential for their operation. However, such localization in large scale networks is challenging. Additionally, the traditional localization techniques based on GPS and satellite navigation provide low accuracy and are useless indoors. Therefore, an alternative technology has to be proposed. This motivates this project’s study of using wireless camera networks to track moving objects. In particular, first standard localization techniques used in a WSN are studied. Then it is explained how to detect and track objects using video cameras. Finally, an analysis is made on how video images and sensor measurements are combined in order to obtain better results. The good performance of these fusion techniques is illustrated with examples. It is demonstrated in a simulation that the localization techniques can perfectly track a moving object if there is no measurement noise. These results also demonstrate the tradeo↵ between l...