Multi-camera collaborative analysis method and system based on topological structure of network system

Abstract

The invention discloses a multi-camera collaborative analysis method and system based on a topological structure of a network system, and the method comprises the steps: obtaining position informationof each camera and a monitoring subnet, and building the map mapping of the monitoring subnet and a global scene; carrying out mathematical abstract modeling on the monitoring subnet and the global scene map mapping to obtain a corresponding topology node connection graph; initializing tracking information, obtaining target information data, and determining the start of tracking; carrying out single-target tracking by adopting a particle filtering algorithm based on scale invariant features; adopting a hierarchical cascade tracking strategy and a joint data association algorithm in the topology subnet to cooperatively control each monitoring camera; searching in the global topological map by adopting a space-time association algorithm; and determining situations such as target tracking termination and loss, and stopping tracking. A suspicious target can be tracked across scenes, the system adopts a distributed hierarchical search method, the tracking cost is reduced, and the suspicious target can be tracked in real time.

Description

technical field [0001] The invention relates to the technical field of visual monitoring, and relates to a camera-assisted tracking method and system. Background technique [0002] With the rapid development of network technology, video security monitoring technology has been popularized and developed on a large scale, and more and more key parts of public activity areas have deployed video surveillance. These surveillance cameras are distributed and operated independently to ensure the public safety of the city and provide auxiliary surveillance information and event traceability credentials. Therefore, analysis of video image abnormal information, real-time tracking of suspicious objects, and improvement of the intelligence of network monitoring make the field of image processing at the forefront of technology and have a broad market demand. [0003] In the scope of current practical applications, most intelligent target tracking methods are only aimed at a single camera ...

AI Technical Summary

Problems solved by technology

[0006] (2) Target occlusion problem: occlusion is a common and difficult problem in target tracking. The lack of target information during the occlusion process will make tracking unstable

[0007] (3) Multi-camera collaborative control problem: It is necessary to determine which camera to use at each moment, that is, data association and information fusion between multiple cameras

Images