Method for detecting abnormal articles in fixed environment in real time

Abstract

The invention discloses a method for detecting abnormal articles in a fixed environment in real time, which can identify normal articles and abnormal articles in the fixed environment. The method comprises the following steps: establishing a normal article database in a fixed environment, detecting all objects in the fixed environment by using a target recognition algorithm, and designing a perceptual hash algorithm to compare detected articles with articles in the normal article database one by one to obtain a similarity value; and when the similarity value is lower than a set threshold value, regarding the object as an abnormal object, and adding the abnormal object into an abnormal object database. The method has the advantages that abnormal objects which should not appear in the sceneoriginally can be detected, manpower is greatly liberated, and the life quality of people is improved. Articles in the environment can be dynamically monitored in real time, early warning is accelerated, and safety performance is improved.

Description

technical field [0001] The invention relates to a real-time monitoring method for abnormal items, and belongs to the technical field of deep reinforcement learning. Background technique [0002] Usually, for a specific environment, the appearance of abnormal items is likely to cause insecurity. For example, objects occupying the fire exits, abnormal objects in subway stations, etc. Real-time detection of abnormal objects has important social significance and economic value. [0003] However, in the existing monitoring system, it mainly relies on the video monitoring system, and there are staff for human eye recognition. Although there are already some anomaly detection and early warning methods for human behavior, various anomalies in subway stations may also be caused by other It can be seen that the existing detection and early warning methods cannot allow the control center to obtain various information intuitively, and at the same time, various emergency plans cannot be...

AI Technical Summary

Problems solved by technology

[0003] However, in the existing monitoring system, it mainly relies on the video monitoring system, and there are staff for human eye recognition. Although there are already some anomaly detection and early warning methods for human behavior, various anomalies in subway stations may also be caused by other The reason is caused; it can be seen that the existing detection and early warning methods cannot allow the control center to obtain various information intuitively, and various emergency plans cannot be rehearsed, and the accuracy of the emergency plan cannot be verified

Images