粒子滤波目标跟踪MATLAB

========================================================================== DISCLAIMER: The demo software is designed in matlab and the result produced by the matlab code is not exactly same as the result presented by the paper since the result in the paper is produced originally with C++. The source code and data is made publicly available mainly for helping those who are interested in the implementation details of BPF. Also for the sake of simplifying the implementation, the mixture particle filter is not used anymore and instead a set of indepedent particle filters is used. However, this does not affect the core of BPF which is the construction of mixture proposals. Lastly, there is no guarantee for the presense of bugs in the code. ========================================================================== This directory contains the most recent version of the public distribution of BPF in Matlab. There are a mex version of adaboost detector and a function of getting adaboost confidence. For copywright issues, please refer to a file called "COPYING.txt" in this directory. There are several subdirectories: BPF: the source code for BPF external: code from external sources that are publicly available - KPMtools by Kevin Murphy [ http://www.cs.ubc.ca/~murphyk/Software/index.html ] - lightspeed by Tom Minka [ http://research.microsoft.com/~minka/software/lightspeed/ ] data: a sequence of image files For the first time user: 1. Start Matlab 2. >startup 3. >test(TrackerBPF) For details, please start reading BPF/@TrackerBPF/demo.m NOTE: For the displaying purpose, color coded square boxes are drawn for tracked targets. However, the color information is only extracted from the black box within a color coded square box. The boxes that are close enough to be interacting are colored in black. Acknowledgements: Thank Wei-Lwun Lu for contributing his implementation of computing hsv color histograms using integral image techniques. Also thank Kevin Murphy and Tom Minka for their matlab tool box. Kenji Okuma okumak[at]cs.ubc.ca