A method for detecting the loss of items in security monitoring

Abstract

The invention provides a method for detecting the loss of items in security monitoring, which specifically includes: acquiring video stream data, initializing fast background B _f , slow background B _s and foreground statistics template F _m , get fast foreground F _f and slow foreground F _s , update fast background B _f and slow background B _s , filter moving targets, update the foreground statistics template F _m , Foreground statistical template F _m Perform binarization to obtain the final result prospect F _result , to the final result prospect F _result Perform morphological operations, corrode and expand to obtain complete targets, track complete targets, judge items left and lost, and update the slow background B of the corresponding target area _s and foreground statistics template F _m and other steps. The detection method for items left behind in security monitoring is based on dual background modeling and foreground statistical templates, which can solve the problem of many false positives and missed negatives caused by poor background modeling in traditional background modeling methods and the need to detect the foreground. The problem of uncertainty in time and prospect duration.

Description

technical field [0001] The invention relates to the technical field of security monitoring, in particular to a method for detecting the loss of items left in security monitoring. Background technique [0002] With the popularization and widespread use of network surveillance cameras, the detection of lost items is widely used in the field of security protection. The detection of lost items is mainly used in public places. It is used to detect whether valuables have been removed and can issue an alarm in time. [0003] The prior art is roughly divided into two types. The first is the method based on background modeling: background modeling is carried out on the monitoring scene through various methods, such as mixed Gaussian model modeling, after background modeling, the foreground is detected, and then the accurate target position is determined through morphological operations. The second is the method of target detection based on convolutional neural network: detect each ...

AI Technical Summary

Problems solved by technology

[0005] Aiming at the deficiencies of the prior art, the present invention proposes a method for detecting items left behind and lost in security monitoring. Based on double-background modeling and foreground statistical templates, it can solve the problems caused by poor background modeling in traditional background modeling methods. issues of positives and false negatives and the uncertainty of the time it takes to detect a foreground and the duration of a foreground

Images