Adaptive foveal sensor for target tracking

A target tracker based on an adaptive foveal sensor and implemented using particle filters is presented. The foveal sensor's field of view includes a high sensitivity "foveal" region surrounded by a lower sensitivity region. The location and acuity of the foveal region are adaptively adjusted, using the estimated target position and spread of particles. A Gaussian resampling method reduces the adverse effects of occasional large observation errors. Monte Carlo simulations show that the foveal sensor with adaptive acuity always outperforms a fixed acuity sensor; also, the particle outperforms previous Kalman filter-based implementations of the non-adaptive foveal sensor.

[1]  L. Li,et al.  Target tracking with an attentive foveal sensor , 2000, Conference Record of the Thirty-Fourth Asilomar Conference on Signals, Systems and Computers (Cat. No.00CH37154).

[2]  Neil J. Gordon,et al.  A tutorial on particle filters for online nonlinear/non-Gaussian Bayesian tracking , 2002, IEEE Trans. Signal Process..

[3]  Nando de Freitas,et al.  Sequential Monte Carlo Methods in Practice , 2001, Statistics for Engineering and Information Science.

[4]  R. Martin,et al.  Nonlinear Filtering Models Of Attentive Vision , 1996, 1996 IEEE International Symposium on Circuits and Systems. Circuits and Systems Connecting the World. ISCAS 96.