Cluster scene intelligent monitoring method and system based on cluster trajectory classification

Abstract

The invention discloses a cluster scene intelligent monitoring method and system based on cluster trajectory classification. The method comprises the steps of first, adopting a key point tracking algorithm for obtaining a motion trajectory of a moving object in a monitoring video; then, adopting a spanning tree algorithm of a global uniform Markov random field model for connecting fragment trajectories with the time-space coherence to obtain a complete trajectory tree; then, adopting a confidence-based correlated topic model for learning middle layer features in the trajectory tree and redescribing the middle layer features; then, adopting a multi-class support vector machine classifier for classifying the trajectory tree on the basis of the trajectory tree described by the middle layer features; finally utilizing the confidence-based correlated topic model and a model of the multi-class support vector machine classifier and adopting a minimization information entropy algorithm for classifying the fragment trajectories. The complete cluster scene intelligent monitoring system is designed to replace manual analysis on cluster scene monitoring videos, manpower resources are saved, and the system robustness is good.

Description

technical field [0001] The invention relates to the field of computer vision, which is based on a tracking algorithm and a topic model. By automatically analyzing the image sequence of the monitoring video recorded by the monitoring equipment, the target in the dynamic scene is positioned, identified and tracked, and based on this Analyze and judge the behavior of the target online, so as to be able to carry out daily management, predict abnormal behavior, and timely alarm and emergency response to abnormal behavior. Compared with the traditional intelligent monitoring system, the present invention can be applied to various scenarios, especially cluster scenarios, and has the advantages of strong intelligence and wide application. Background technique [0002] With the development of society, all kinds of terrorist attacks and dangerous and violent crimes are increasingly threatening the safety of national life and property. Effectively combating various crimes and effective...

AI Technical Summary

Problems solved by technology

[0004] At present, with the advancement of network technology, the transmission and storage of surveillance video have been greatly improved. However, because the intelligent analysis of surveillance video has not been effectively solved, it has caused a bottleneck in the development of surveillance systems.

Images