TARGET TRACKING METHOD AND TRACKING SYSTEM MEASURING AND DRIVING BY PROPAGATING MARGINAL DISTRIBUTION

摘要

The present invention relates to the field of multi-sensor information integration, and provides a target tracking method and tracking system measuring and driving by propagating a marginal distribution, comprising: according to a previous moment marginal distribution and probability of existence thereof, obtaining a current moment predicted marginal distribution and probability of existence thereof; according to the predicted marginal distribution and probability of existence thereof, and using a Bayesian rule, sequentially processing measurement data of the current moment to obtain an updated marginal distribution and probability of existence thereof; generating, using the current moment measurement data, a current moment marginal distribution of a new target, and specifying a probability of existence therefor; combining the same respectively with the updated marginal distribution and the probability of existence thereof to generate current moment marginal distributions and probabilities of existence thereof; removing, from the marginal distributions of the current moment, marginal distributions having a probability of existence less than a first threshold, the removed marginal distributions and probabilities of existence thereof serving as a recursive input for a subsequent moment; and simultaneously extracting marginal distributions having a probability of existence greater than a second threshold as a current moment output.