Fast Image-Based Tracking by Selective Pixel Integration

Abstract

We provide a fast algorithm to perform image-based tracking, which relies on the selective integration of a small subset of pixels that contain a lot of information about the state variables to be estimated. The resulting dramatic decrease in the number of pixels to process results in a substantial speedup of the basic tracking algorithm. We have used this new method within a surveillance application, where it will enable new capabilities of the system, i.e. real-time, dynamic background subtraction from a panning and tilting camera. 1. Introduction/Philosophical Approach One of the fundamental tasks of real-time processing has been image-based tracking through a video sequence using a parametric motion model. This includes both tracking of a moving object through an image sequence [5] as well as registering of whole images to a reference view to provide software video stabilization [2,8]. Figure 1. Selective pixel integration selects pixels with a high information content with respe...