Fast Graph Segmentation Based on Statistical Aggregation Phenomena

Object tracking in computer vision refers to the task of tracking individual moving objects accurately from one frame to another in an image sequence. Several tracking methods have been proposed in the recent literature capable of coping with a certain degree of occlusions of the objects. However, no comparative analysis of such methods has been presented to date and both the expert and the newcomer to this area may be confused about the relative effectiveness of each method when compared under the same level of complexity of the dynamic scene. In order to fulfill this need, this paper proposes a set of analysis criteria and provides a comparative review of the main recent tracking methods, in particular with respect to their capability of tracking objects under occlusions.