A Space Registration Algorithm for Multi-Sensor Target Tracking Using an Entropy Function
Guo Jun-ju · Xi'an Jiaotong Daxue xuebao · 2014
An algorithm using a mini-max entropy function is proposed for sensor registration.The algorithm uses an entropy function as the optimization criterion,and the objective function is formulated based on a measurement model.The principle of the maximum entropy is employed to transform the absolute value of the objective function into a maximum entropy function,and then aquasi-Newton method is used to minimize the maximum function to obtain the sensor biases.Simulation results in both the scenarios of a single target and the multiple targets and comparisons with the traditional registration algorithm show that the method has a significant an improvement in performance of reducing errors of the sensor biases.