Data association method based on Gaussian mixture model

Abstract

The invention discloses a data association method based on a Gaussian mixture model. The method comprises the following steps of obtaining a state model of multi-target tracking according to the initial states of the targets and a first preset matrix; obtaining an observation model of the targets according to the state model and a second preset matrix, and calculating the predicted position and the innovation covariance of the targets according to the observation model; obtaining a plurality of target effective measurements according to the observation model and a preset wave gate, and constructing a Gaussian mixture model by the plurality of target effective measurements; establishing a posterior probability model according to the Gaussian mixture model, updating parameters in the posterior probability model by using a maximum expectation method, and obtaining a target posterior probability according to the parameters; and constructing an incidence matrix according to the target posterior probability, and searching the incidence matrix to obtain the targets and the target effective measurements. According to the invention, the Gaussian mixture model is used for multi-target tracking, complex matrix splitting of a joint probability data association method is avoided, the calculated amount is reduced, the tracking precision is high, and the method is suitable for real-time multi-target tracking.

Description

technical field [0001] The invention belongs to the technical field of multi-target tracking, and in particular relates to a data association method based on a Gaussian mixture model. Background technique [0002] With the development of radar technology, multi-target tracking has become more and more widely used in military and civilian fields. Data association is the core part of multi-target tracking. The essence of data association is to associate measurements with existing tracks. [0003] In recent years, many data association methods have been proposed, such as the nearest neighbor data association method, the probabilistic data association method, and the joint probability data association method. Among them, the nearest neighbor data association method selects the measurement closest to the target prediction position for state update; The method considers that all measurements in the tracking gate may come from the target, but the probabilities of the measurement fr...

AI Technical Summary

Problems solved by technology

[0004] However, when the above method is used for multi-target tracking, there is a contradiction between the tracking accuracy and the calculation amount, that is, the calculation amount is reduced, and the tracking effect is not good, but the tracking effect is good, and the calculation amount explodes.

Images