A Probabilistic Nearest Neighbor Multi-Target Tracking Method Based on Fuzzy Clustering

Abstract

The invention relates to a probability nearest neighbor domain multi-target tracking method based on fuzzy clustering. The method comprises the steps that the possibility situation from a real target is measured in a nearest neighbor domain wave door firstly, the association degree distinguishing standard of effective measurement and existing flight paths is improved based on a fuzzy clustering theory, and a target state estimation and covariance updating equation is perfected; meanwhile, a distributed parallel processing structure is adopted, flight path fusion and state estimation are carried out on sub-flight-path information output by sub-sensors, tracking real-time performance is guaranteed, meanwhile, the robustness of a system is enhanced, and tracking precision is improved. It is indicated by an experimental result that in the multi-target tracking system with a radar / infrared multisensor fused under a clutter environment, compared with a nearest neighbor domain standard filter method, the tracking effect is good, and the method is suitable for tracking multiple maneuvering targets under the clutter environment.

Description

technical field [0001] The invention belongs to the field of multi-sensor multi-target tracking, in particular to a probabilistic nearest neighbor multi-target tracking method based on fuzzy clustering. Background technique [0002] The core part of multi-sensor multi-target tracking problem is data association and state estimation. For a typical multi-sensor multi-target tracking system composed of 3D radar and infrared, due to the many uncertain interference factors in the sensor observation process and target tracking environment, the data association and target state estimation problems are complex and difficult, especially when the target maneuvers or the distance is relatively close When , it is easy to cause ambiguity in the association of multi-sensor and multi-target data, which will affect the tracking performance. Therefore, research on multi-target tracking technology in clutter environment has important application value. [0003] The existing algorithms for d...

AI Technical Summary

Problems solved by technology

Most of the traditional data association algorithms are less reliable when the association is ambiguous, which can easily lead to a significant decrease in the target tracking accuracy; and when the number of targets increases, the calculation amount of the traditional data association algorithm increases sharply, which limits its application range

Images