Preliminary studies on the taxonomy of object's tracking algorithms in video sequences

Different techniques for tracking objects in controlled environments using video cameras have been proposed. These state of the art algorithms are focused especially on how to find a better segmentation of the tracking object and also on how to make this segmentation stable through time, regardless of temporal changes on the morphology of the object. Unlike any of that, this article reviews the state of the art, focusing on algorithms for segmentation of the scene and of tracking objects, then addresses the previous steps in the creation of a binary image that segments the objects and convert them into useful data, found frame by frame to be used afterwards for tracking. The intention is to classify the methods of temporal matching between the binary images which are the outcome of the segmentation of foreground and background into general groups, in order to give an organized starting point to the advances made regarding the tracking of moving objects with fixed cameras and to be able to adapt faster to the implementation of tracking on the new advances in specific techniques in the field of the proposed taxonomy.