Object Tracking And Detection PDF PPT Presentation And Seminar Report
A method of following an object through successive image frames to determine its relative movement with respect to other objects.
In a tracking scenario, an object can be defined as anything that is of interest for further analysis. Objects can be represented by their shapes.
In this paper we describe a method for tracking objects whose number is unknown and varies during tracking. Based on preliminary results of object detection in each image which may have missing and/or false detection, the multiple object tracking method keeps a graph structure where it maintains multiple hypotheses about the number and the trajectories of the objects in the video. The image information drives the process of extending and pruning the graph, and determines the best hypothesis to explain the video. While the image-based object detection makes a local decision, the tracking process con?rms and validates the detection through time, therefore, it can be regarded as temporal detection which makes a global decision across time.
The multiple object tracking method gives feedbacks which are predictions of object locations to the object de-tection module. Therefore, the method integrates object detection and tracking tightly. The most possible hypothe-sis provides the multiple object tracking result. The exper- imental results are presented.