A template updating reasoning engine which can deal with fundamental constraints on the spatial-temporal continuity of target's motion is proposed. By analyzing target's continuously adaptive distributions image, a voting method can estimate the tracking window's scale. In updating phase, by making further computation of likelihood of target model and candidate model, both the model and scale can be automatically updated in time. The tracking ability of KBT can be improved. © 2011 Springer-Verlag.
CITATION STYLE
Han, R. (2011). Kernel based visual tracking with reasoning about adaptive distribution image. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 7004 LNAI, pp. 529–536). https://doi.org/10.1007/978-3-642-23896-3_65
Mendeley helps you to discover research relevant for your work.