Sequential Kernel Density Approximation and Its Application to Real-Time Visual Tracking

• Published in: IEEE transactions on pattern analysis and machine intelligence Vol. 30; no. 7; pp. 1186 - 1197
• Main Authors: Bohyung Han, Comaniciu, D., Ying Zhu, Davis, L.S.
• Format: Journal Article
• Language: English
• Published: Los Alamitos, CA IEEE 01.07.2008; IEEE Computer Society; The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
• Subjects: Algorithms; Application software; Applied sciences; Approximation; Approximation algorithms; Artificial Intelligence; Computational modeling; Computer science; control theory; systems; Computer Simulation; Computer Systems; Computer vision; Density; Density functional theory; Exact sciences and technology; Gaussian processes; Image Enhancement - methods; Image Interpretation, Computer-Assisted - methods; Kernel; Kernels; Mathematical analysis; Mathematical models; Models, Statistical; Motion; Pattern Recognition, Automated - methods; Pattern recognition. Digital image processing. Computational geometry; Probability density function; Representations; Reproducibility of Results; Sensitivity and Specificity; Statistical; Statistical Distributions; Studies; Subtraction Technique; Target tracking; Tracking; Videos; Visual; Statistical; Computer vision; Tracking; Mixed distribution; Density estimation; Target tracking; Density measurement; mean shift; real- time computer vision; Motion estimation; Computer simulation; Image processing; Compactness; Video signal; mode propagation; Real time; Modeling; Kernel method; Gaussian process; Probability density function; object tracking; Density function; online target appearance modeling; Pattern analysis; Artificial intelligence; Kernel density approximation
• ISSN: 0162-8828; 1939-3539
• DOI: 10.1109/TPAMI.2007.70771

Visual features are commonly modeled with probability density functions in computer vision problems, but current methods such as a mixture of Gaussians and kernel density estimation suffer from either the lack of flexibility by fixing or limiting the number of Gaussian components in the mixture or large memory requirement by maintaining a nonparametric representation of the density. These problems are aggravated in real-time computer vision applications since density functions are required to be updated as new data becomes available. We present a novel kernel density approximation technique based on the mean-shift mode finding algorithm and describe an efficient method to sequentially propagate the density modes over time. Although the proposed density representation is memory efficient, which is typical for mixture densities, it inherits the flexibility of nonparametric methods by allowing the number of components to be variable. The accuracy and compactness of the sequential kernel density approximation technique is illustrated by both simulations and experiments. Sequential kernel density approximation is applied to online target appearance modeling for visual tracking, and its performance is demonstrated on a variety of videos.