Outdoor child unmanned accompanying detection method and device based on community cerebellum monitoring

Abstract

The invention discloses an outdoor child unmanned accompanying detection method based on community cerebellum monitoring. The method comprises the following steps: 1) inputting a community cerebellum monitoring video; 2) judging whether each detected person is a child or an adult according to a pre-calibrated long-shot close-shot frame of the video monitoring scene; 3) extracting the detected REID feature value of each person, and recording the REID feature value in a human body ID historical list; 4) analyzing and determining whether the child is in an unattended state or not according to the adjoint relationship; 5) according to the duration of the unattended state of the child, through a time risk function, calculating to obtain a time risk score; 6) calculating a spatial risk score according to the child activity track; 7) calculating a posture risk score according to the posture recognition of the child; and 8) calculating an overall risk index by adopting a comprehensive scoring method. Through the comprehensive score judgment of the activity time, the activity track and the posture state of the unattended child, the alarm information of the unattended child is generated, and the face or human shape information does not need to be input in advance.

Description

technical field [0001] The invention relates to the technical field of detection methods for unaccompanied or lost children, in particular to an outdoor unaccompanied detection method and device based on community cerebellum monitoring. Background technique [0002] Existing detection methods for unaccompanied or lost children mainly include solutions based on positioning devices, face recognition, and human body recognition. Most of the detection of lost children or the elderly similar to the present invention is based on the detection of positioning equipment, or pre-input people. Unaccompanied police with face information, or use face or human body search to find out where children or elderly people have been after a missing incident occurs, but it does not take into account that in most outdoor scenes, lost children do not wear positioning equipment and do not record in advance Face information, and it is necessary to first realize the detection of unaccompanied situatio...

AI Technical Summary

Problems solved by technology

[0012] Aiming at the deficiencies of the prior art, the present invention provides a method and device for unaccompanied detection of outdoor children based on community cerebellar monitoring, which has the advantages of easy use and solves the problem of the existing detection method of outdoor unaccompanied children based on community cerebellar monitoring and inconvenient use of the device

Images