Research On The Detection Method Of Fence Compliance Based On Open CV

Power engineering is an important lifeline to ensure the rapid and stable development of my country’s economy.The significance of the research on the compliance placement of construction safety fences is not only to ensure the personal safety of construction personnel and reduce the potential safety hazards in the construction process,but also to ensure the smooth progress of power system engineering.Today,with the continuous improvement of power system engineering management and control,a large number of visualization applications are combined with power system engineering management and control,and there are massive construction site images and video information.However,in practical applications,most of the problems in the compliance of safety fences on construction sites still need to be discovered by engineering supervision sites or video surveillance.Information will be wasted.This paper takes ensuring the normative placement of safety fences at power system construction sites as the starting point,and takes full use of the picture and video information of power system engineering visualization applications as the breakthrough point,combined with the actual project,to carry out the Open CV-based fence compliance placement detection method Research.Through the analysis and summary of the fence images of the power system construction site,and at the same time,a large number of image processing algorithms are carried out to try and analyze the results.It is proposed to combine two parts of fence extraction and fence gap existence judgment based on Open CV and fence gap classifier training based on Open CV to jointly realize fence gap compliance detection.The main work of this paper is as follows:Aiming at the problem of fence extraction and fence gap judgment,the principles of fence compliance judgment and image shooting were determined,and a research on the algorithm process design of fence extraction and fen ce gap judgment was carried out.By trying the current popular related image processing algorithms and a large number of experiments,arranging the image processing operations and changing the content of different image processing links in combination with related problems,Finally,the related problems are solved.A set of algorithm flow of fence extraction and fence gap judgment is obtained,which takes the gray value of the target pixel area of the image as the starting point and uses the connected area to realize the label distinction as a springboard.Aiming at the problem of fence gap detection,this paper trains a feature classifier dedicated to identifying fence gaps.By constructing a targeted sample set to further clarify the gap characteristics,In order to improve the recognition accuracy,a classifier training process implemented by Haar feature + Ada Boost + Cascade cascade is finally constructed.And through this process,a Haar feature classifier dedicated to fence gap detection is obtained.By calling the classifier and combining with the actual experiment,the sample set is optimized and reconstructed to further enhance the recognition accuracy.According to the actual use requirements and placement status of safety fences in power engineering,and in response to the call to improve the level of engineering management and control,combined with the video image information of visual applications,This paper proposes an Open CV-based fence placement compliance detection method.Aiming at the placement of safety fences in power construction,it uses visual terminal equipment combined with engineering to collect on-site information,and process,detect and count on-site information.The experimental results show that the proposed method can correctly realize the compliance judgment of fence placement,which has very precise practical significance in engineering safety application.