On-line selection of discriminative tracking features：（在线的选择有识别力的跟踪特性）.pdfVIP

On-Line Selection of Discriminative Tracking Features ∗ Robert T. Collins 1, Yanxi Liu 2 and Marius Leordeanu 2 1 The Pennsylvania State University 2 Carnegie Mellon University Abstract This paper presents an on-line feature selection mechanismfor evaluating multiple features while tracking and adjusting the set of features used to improve tracking performance. Our hypothesis is that the features that best discriminate between object and background are also best for track- ing the object. Given a set of seed features, we compute log likelihood ratios of class conditional sample densities from object and background to form a new set of candidate features tailored to the local object/background discrimination task. The two-class variance ratio is used to rank these new features according to how well they separate sample distributions of object and background pixels. This feature evaluation mechanism is embedded in a m ean-shift tracking system that adap- tively selects the top-ranked discriminative features for tracking. Examples are presented that demonstrate how this method adapts to changing appearances of both tracked object and scene background. We note susceptibility of the variance ratio feature selection method to distraction by spatially correlated background clutter, and develop an additional approach that seeks to minimize the likelihood of distraction. Keywords: computer vision, tracking, time-varying imagery, feature creation, feature evaluation and selection. ∗This work is supported by DARPA/IAO HumanID contract N00014-00-1-0915, by DARPA/IPTO MARS2020 contract NBCHC020090 and by DARPA/IXO VIVID contract NBCH1030013. 1. Introduction Two decades of vision research have yielded an arsenal of powerful algorithms for object track- ing. Multiple moving objects can be effectively tracked in