Electronic fence risk early warning method based on deep learning

Abstract

The invention relates to an electronic fence risk early warning method based on deep learning, and the method comprises the steps of constructing a conical barrel detection model based on deep learning, carrying out the training, recognizing a conical barrel in an image collected in real time through the conical barrel detection model after the training is completed, selecting a recognition point after the recognition is completed, carrying out the calculation and determination of an electronic fence range, and further identifying a moving target in the image, judging whether the moving target is within the coverage range of the electronic fence, and if so, sending out an alarm signal. According to the invention, it can be judged that an electronic fence is established at an accident occurrence site through image recognition, and whether an obstacle hindering accident investigation exists in the range of the electronic fence is judged, so that the influence on accident elimination is avoided.

Description

technical field [0001] The invention relates to the technical field of deep learning, in particular to a method for early warning of electronic fence risks based on deep learning. Background technique [0002] The electronic fence is currently the most advanced perimeter anti-theft alarm system. Usually, the electronic fence is set outdoors and installed along the original wall (such as a brick wall, cement wall, iron fence or tapered barrel), and the alarm signal is transmitted through the signal transmission device. It is sent to the back-end control center to display the working status of the defense zone. Most electronic fencing today is used for agricultural fencing and other forms of animal management, but is also often used to enhance security in sensitive areas such as military installations, prisons and other security-sensitive locations, where lethal voltages are present. [0003] At present, most of the perimeter alarm systems still stay on the front-end alarm an...

AI Technical Summary

Problems solved by technology

[0003] At present, most of the perimeter alarm systems still stay on the front-end alarm and the control and management of a single device. If you want to achieve cross-regional remote control and management, you must rely on a third-party alarm system for integration

However, manufacturers that provide alarm systems may not be able to provide open protocols and interfaces, and it is difficult to form a unified integrated platform, resulting in a "decentralized management" system for the security system that should have been planned in a unified manner. At the same time, the existing electronic fence is inconvenient Difficult to change and adjust in real time according to demand

Images