Localizacion y seguimiento de objetos en videos distorsionados.

Video trackers defined as the analysis of video sequences for the purpose of establishing the location of a target over a sequence of frames have been rapidly evolving in the last two decades. Despite this effort, practical video trackers deployed today are limited by video distortions generated by compression and transmission of videos. These distortions hampers their performance in following objects throught complex scenes. The main objetive of this proposal is to design and implement robust algorithms to localize and track objects. This goal will be achieved by adding visual perceptual features to the trackers to make them robust with respect to distortion. The first specific objective is to analyze the impact of distortions caused by compression and packet losses on the performance of trackers. Second, we develop robust trackers with respect to distortions generated by compression and packet loss. Finally, we evaluate the performance of this tracker in a public dataset with distorted videos. The expected outcomes are a robust tracker with respect to video distortions and the training of an undergraduate student in this topic.