Security alarm method and system based on video recognition and storage medium

Abstract

The invention discloses a security alarm method based on video recognition. The method comprises an acquisition step of acquiring a video captured by a camera, wherein the video comprises multiple frames of video images, a first marking step of performing feature recognition on target objects in the video images and marking a target object which is not recognized to obtain a marker, and an early warning step of judging whether the distribution density change of the markers of adjacent frame video images exceeds a first preset value and carrying out early warning if so. The invention also provides a security alarm system based on video recognition and a computer-readable storage medium. The invention provides the method, the distribution density change of the markers in a recognition area is judged to judge the presence or not of invasive alien species, a technical problem of judging all invasive alien species can be covered, and the cost of the technical implementation is low.

Description

Technical field [0001] The invention relates to the technical field of image recognition, in particular to a security alarm method, electronic equipment and storage medium based on video recognition. Background technique [0002] At present, poultry farms are generally set up in places far away from villages, towns and urban areas or close to mountains and forests. They rely on manual patrols and fences by breeders for anti-theft safety protection. There are also management methods that rely on camera devices for monitoring. Regularly monitor the video to view the security work, or rely on the existing video security monitoring technology to monitor the intrusion of unfamiliar and suspicious persons or species and carry out relevant early warning to implement security measures. The traditional methods of manual patrol or manual video monitoring consume too much energy and time. The existing video monitoring and early warning security technology needs to judge the invasion of vari...

AI Technical Summary

Problems solved by technology

[0002] At present, poultry farms are generally set up in remote villages, towns, urban areas or places close to mountains and forests, relying on manual inspections and fences by farmers for anti-theft security protection, and there are also management methods that rely on camera devices for monitoring, but most of them rely on manual labor. Regularly monitor the video to check the security work, or rely on the existing video security monitoring technology to monitor the intrusion of strange and suspicious people or species, and then carry out relevant early warning to implement security measures

The traditional methods of manual patrol or manual video monitoring are too labor-intensive and time-consuming, while the existing video monitoring and early warning security technology needs to judge the invasion of various species, and the implementation method needs to fully cover all invasive species that may attack the poultry. It has certain technical difficulties and high implementation cost

Images