Extended target tracking method based on cooperative target information

Abstract

The invention discloses an extended target tracking method based on cooperative target information. According to the method, a cooperative target periodically detects the position and posture of the cooperative target itself, and publishes the identity, position and posture information of the cooperative target through a communication network; and an observation station receives the information sent by the cooperative target, fuses the information with the target information detected by an observation station sensor and performs tracking of a plurality of extended targets, wherein the target tracking process comprises: firstly, describing motion of the targets by utilizing a group of motion models; constructing a filter for each motion model; screening the measurement values of the sensorthrough a double tracking gate according to the position and attitude information of the extended targets, and reconstructing a one-step prediction value for the screened measurement values by using the identity information and attitude features of the cooperative target to perform data association and filtering; and finally, carrying out fusion according to the probability that the targets are indifferent models, and taking the probability weighted sum of filtering estimation values of all filters as target state estimation. The method can effectively improve the tracking precision of the extended targets in the presence of the cooperative target.

Description

technical field [0001] The invention belongs to computer and automation technology, and mainly relates to the technical field of target tracking, in particular to an extended target tracking method based on cooperative target information. Background technique [0002] Target tracking technology has very important application value in both military and civilian fields. This technology is the key development direction of industry and academia. [0003] Chinese patent application: Intelligent vehicle target tracking system and method based on fusion of DSRC and vehicle sensors (application number: CN201610126332.3), which discloses a vehicle-to-vehicle communication method to obtain the motion state and identity information of the tracked target for target tracking method, which only improves the problem that the traditional method does not make full use of the identity information and attitude information issued by the target vehicle. First of all, it does not consider the pr...

AI Technical Summary

Problems solved by technology

[0003] Chinese patent application: Intelligent vehicle target tracking system and method based on fusion of DSRC and vehicle sensors (application number: CN201610126332.3), which discloses a vehicle-to-vehicle communication method to obtain the motion state and identity information of the tracked target for target tracking method, this method only improves the problem that the traditional method does not make full use of the identity information and attitude information released by the target vehicle

First of all, it does not consider the problem that the motion model of the tracked target is relatively simple, and it cannot track the target with a complex motion state; secondly, it does not consider the generality of the method, and the application scenario is relatively single

Chinese patent application: An interactive multi-model target tracking method with optimal motion mode switching parameters (application number: CN201410234810.3), which discloses an interactive method using the linear minimum variance theory to derive the optimal motion mode switching parameters A multi-model target tracking method, which takes into account the correlation between motion modes, uses all the covariance information between motion modes, avoids the problem of missing correlation in traditional methods, but this method does not consider target tracking Even if there is only one target in the scene, measurement noise and clutter problems will be introduced. The data association method is not used to establish the corresponding relationship between the target and the measurement. This method also does not consider the data association of the extended target with gesture features.

Images