On-line bias compensation for dissimilar sensor fusion tracking
异类传感器融合跟踪系统配准偏差的在线补偿

For the time-variant registration errors of multi-sensor fusion tracking system,an on-line bias estimation and compensation algorithm for multi-sensor fusion tracking is proposed.A dynamic model of the bias is formulated based on the target information of measurements and estimation.A minimization of the likelihood function and Kalman filter are used to estimate the bias on-line,and the measurements are compensated for estimated bias.Finally,an example for the estimation of the bias parameters of radar and IRST sensors is given to illustrate the effectiveness of the approach.