Gaussian mixture cphd filtering method with track correlation and extraction capabilities

Abstract

The invention relates to the technical field of target tracking and specifically relates to a Gaussian mixture CPHD (Cardinalized Probability Hypothesis Density) filtering method with flight track association and extraction capacity. The invention aims at providing the CPHD filtering method with accurate and efficient state estimation and flight track maintaining capacity. The method is realized through adoption of a Gaussian mixture technology. According to a new filter, problems existing in a CPHD filter are solved; left PHD can be accurately allocated; accurate target state estimation can be carried out; a tracking effect is improved clearly; and a target flight track can be maintained through utilization of a flight track association and extraction process independent of a filtering process. The flight path maintaining and filtering processes are non-interfering. Operation load of the filtering process is not increased, efficient operation of the filtering process is ensured, and the target track also can be extracted flexibly offline.

Description

technical field [0001] The present invention relates to the technical field of target tracking, in particular to a Gaussian mixture collection potential probability hypothesis density (Cardinalized Probability Hypothesis Density, CPHD) filtering method with track association and extraction capabilities. Background technique [0002] With the development of modern technology, the scene of target surveillance is becoming more and more complex, and the requirements for target surveillance capabilities are also getting higher and higher. In the current target tracking scenario, sensors often need to track a large number of densely distributed multiple targets, which leads to complex data association, increased calculation, increased target loss rate, and unstable estimation of the number of targets and target states. , which poses a great challenge to the data processing capability of the sensor. In the face of such a scenario, traditional multi-target tracking methods are diff...

AI Technical Summary

Problems solved by technology

However, the CPHD filter also has some problems, such as the incorrect allocation of the PHD of the missed target, the difficulty of estimating the target state, and the lack of track maintenance capabilities, etc., which makes the CPHD filter still have targets in the face of more complex tracking scenarios. The number and target state estimation are unstable and the accuracy is not high, so the tracking performance of CPHD still cannot meet the current increasingly complex target monitoring needs

Images