Door switch state recognition method based on image processing

Abstract

The invention discloses a door switch state recognition method based on image processing. Firstly, it is judged whether the door state changes within a certain range of frames. If there is a change, a multi-frame method is used for monitoring; When the multi-frame method is used for monitoring, the door switch state detection is based on the relevant pixel information of the door area in the video stream, by comparing the difference between the current image and the reference image, and by comparing the motion between the current image and the preceding and following frame images Changes that come out of the door in the current state. Based on single-frame comparison recognition, the state of the door is recognized by using prior knowledge. The accuracy and anti-interference performance of the present invention are improved, and multiple algorithms, multiple dimensions, and strong constraints are adopted to limit the accuracy of the results.

Description

technical field [0001] The application belongs to the field of smart security, and relates to a method for accurately identifying the switch state of a door based on image processing. Background technique [0002] my country's intelligence is developing rapidly, and people's demand for smart security is increasing rapidly. As an essential part of various scenarios, the door's correct switch state recognition is directly related to the subsequent security content such as face recognition. As the first process of intelligent security, it has attracted more and more attention. [0003] However, in most of the state detection at present, the way of identifying markers is often used to output status results. Under the influence of this process, the operation of the algorithm will greatly depend on the manual maintenance of markers, especially When the number of detections in the system is large and the markers are easily affected by natural factors, the labor cost is greatly inc...

AI Technical Summary

Problems solved by technology

[0003] However, in most of the state detection at present, the way of identifying markers is often used to output status results. Under the influence of this process, the operation of the algorithm will greatly depend on the manual maintenance of markers, especially When the system has a large number of detections and the markers are easily affected by natural factors, the labor cost is greatly increased, and it also deviates from the automatic identification purpose of the system

In addition, due to the influence of factors such as light / occlusion during monitoring, only using markers for judgment cannot cope with changes brought about by time changes

Images