Distributed target tracking method suitable under camera network

Abstract

The invention relates to a distributed target tracking method suitable under a camera network, and belongs to the technical field of the camera network, distributed processing and target tracking application. The method includes the steps of performing motion prediction and estimation for a target monitored by each camera using square root cubature information filtering; performing information interaction among the cameras through communication, and then performing distributed data fusion of the information using a mean consistency method; and finally, obtaining a stable tracking result through multiple iterations so as to achieve target tracking under the camera network. Compared with the prior art, the method of the invention takes relevant noise between process noise and measurement noise into consideration so that a system with the presence of the relevant noise can also achieve the application of target tracking; the use of the square root cubature information filtering allows the system to avoid the falling into the finite word-length problem of a processor, and meanwhile, the combination of the weighted mean consistency method allows the application in a distributed environment by the system; and the robustness of the method is enhanced while the tracking accuracy is improved.

Description

technical field [0001] The invention belongs to the technical field of camera network, distributed processing and target tracking application, and relates to a distributed target tracking method applicable to the camera network. Background technique [0002] At present, cameras are widely used in security, monitoring and other fields, but the vast majority of applications adopt centralized management methods. However, each camera in the camera network is distributed in different positions in the observation area. If centralized management is adopted, there will inevitably be Wiring is difficult. In addition, the centralized management method will cause a large computing load and storage pressure on the central node. At present, the rise of the Internet of Things and cloud computing has led to the rapid development of distributed processing technology, and its technical advantages have also been highlighted. Therefore, the present invention uses a distributed processing meth...

AI Technical Summary

Problems solved by technology

[0004] The purpose of the invention is to realize the target tracking in the camera network, and solve the problem that the existing target tracking method cannot be applied to large-scale systems due to the use of centralized processing methods, and overcome the existing methods that often do not consider process noise In order to solve the problem of correlation with observation noise, a target tracking method suitable for distributed camera networks is proposed, which can realize target tracking in camera networks and improve target tracking accuracy. It has great potential in the fields of distributed systems and target tracking important role

Images